Truncate web logs to 100K characters (#111)

* set log limit to 100K in browser

.github/workflows/closeinactive.yml

@@ -0,0 +1,23 @@
+# https://docs.github.com/en/actions/use-cases-and-examples/project-management/closing-inactive-issues
+name: Close inactive issues
+on:
+  schedule:
+    - cron: "32 1 * * *"
+
+jobs:
+  close-issues:
+    runs-on: ubuntu-latest
+    permissions:
+      issues: write
+      pull-requests: write
+    steps:
+      - uses: actions/stale@v9
+        with:
+          days-before-issue-stale: 14
+          days-before-issue-close: 14
+          stale-issue-label: "stale"
+          stale-issue-message: "This issue is stale because it has been open for 2 weeks with no activity."
+          close-issue-message: "This issue was closed because it has been inactive for 2 weeks since being marked as stale."
+          days-before-pr-stale: -1
+          days-before-pr-close: -1
+          repo-token: ${{ secrets.GITHUB_TOKEN }}

.github/workflows/containers.yml

@@ -16,6 +16,7 @@ jobs:
     strategy:
       matrix:
         platform: [intel, cuda, vulkan, cpu, musa]
+      fail-fast: false
     steps:
       - name: Checkout code
         uses: actions/checkout@v4

.github/workflows/go-ci.yml

@@ -0,0 +1,32 @@
+# This workflow will build a golang project
+
+name: CI
+
+on:
+  push:
+    branches: [ "main" ]
+
+  pull_request:
+    branches: [ "main" ]
+
+  # Allows manual triggering of the workflow
+  workflow_dispatch:
+
+jobs:
+
+  run-tests:
+    runs-on: ubuntu-latest
+    steps:
+    - uses: actions/checkout@v4
+
+    - name: Set up Go
+      uses: actions/setup-go@v4
+      with:
+        go-version: '1.23'
+
+    # necessary for testing proxy/Process swapping
+    - name: Create simple-responder
+      run: make simple-responder
+
+    - name: Test all
+      run: make test-all

.goreleaser.yaml

@@ -15,4 +15,16 @@ builds:
       - goos: freebsd
         goarch: arm64
       - goos: windows
-        goarch: arm64
+        goarch: arm64
+
+# use zip format for windows
+archives:
+  - id: default
+    format: tar.gz
+    name_template: "{{ .ProjectName }}_{{ .Version }}_{{ .Os }}_{{ .Arch }}"
+    builds_info:
+      group: root
+      owner: root
+    format_overrides:
+      - goos: windows
+        format: zip

README.md

@@ -1,4 +1,7 @@
-![llama-swap header image](header.jpeg)
+![llama-swap header image](header2.png)
+![GitHub Downloads (all assets, all releases)](https://img.shields.io/github/downloads/mostlygeek/llama-swap/total)
+![GitHub Actions Workflow Status](https://img.shields.io/github/actions/workflow/status/mostlygeek/llama-swap/go-ci.yml)
+![GitHub Repo stars](https://img.shields.io/github/stars/mostlygeek/llama-swap)
 
 # llama-swap
 
@@ -11,26 +14,29 @@ Written in golang, it is very easy to install (single binary with no dependancie
 - ✅ Easy to deploy: single binary with no dependencies
 - ✅ Easy to config: single yaml file
 - ✅ On-demand model switching
-- ✅ Full control over server settings per model
 - ✅ OpenAI API supported endpoints:
   - `v1/completions`
   - `v1/chat/completions`
   - `v1/embeddings`
   - `v1/rerank`
   - `v1/audio/speech` ([#36](https://github.com/mostlygeek/llama-swap/issues/36))
-- ✅ Run multiple models at once with `profiles` ([docs](https://github.com/mostlygeek/llama-swap/issues/53#issuecomment-2660761741))
-- ✅ Remote log monitoring at `/log`
-- ✅ Direct access to upstream HTTP server via `/upstream/:model_id` ([demo](https://github.com/mostlygeek/llama-swap/pull/31))
-- ✅ Manually unload models via `/unload` endpoint ([#58](https://github.com/mostlygeek/llama-swap/issues/58))
+  - `v1/audio/transcriptions` ([docs](https://github.com/mostlygeek/llama-swap/issues/41#issuecomment-2722637867))
+- ✅ llama-swap custom API endpoints
+  - `/log` - remote log monitoring
+  - `/upstream/:model_id` - direct access to upstream HTTP server ([demo](https://github.com/mostlygeek/llama-swap/pull/31))
+  - `/unload` - manually unload running models ([#58](https://github.com/mostlygeek/llama-swap/issues/58))
+  - `/running` - list currently running models ([#61](https://github.com/mostlygeek/llama-swap/issues/61))
+- ✅ Run multiple models at once with `Groups` ([#107](https://github.com/mostlygeek/llama-swap/issues/107))
 - ✅ Automatic unloading of models after timeout by setting a `ttl`
 - ✅ Use any local OpenAI compatible server (llama.cpp, vllm, tabbyAPI, etc)
 - ✅ Docker and Podman support
+- ✅ Full control over server settings per model
 
 ## How does llama-swap work?
 
 When a request is made to an OpenAI compatible endpoint, lama-swap will extract the `model` value and load the appropriate server configuration to serve it. If the wrong upstream server is running, it will be replaced with the correct one. This is where the "swap" part comes in. The upstream server is automatically swapped to the correct one to serve the request.
 
-In the most basic configuration llama-swap handles one model at a time. For more advanced use cases, the `profiles` feature can load multiple models at the same time. You have complete control over how your system resources are used.
+In the most basic configuration llama-swap handles one model at a time. For more advanced use cases, the `groups` feature allows multiple models to be loaded at the same time. You have complete control over how your system resources are used.
 
 ## config.yaml
 
@@ -61,13 +67,20 @@ models:
 # Default (and minimum) is 15 seconds
 healthCheckTimeout: 60
 
-# Write HTTP logs (useful for troubleshooting), defaults to false
-logRequests: true
+# Valid log levels: debug, info (default), warn, error
+logLevel: info
 
 # define valid model values and the upstream server start
 models:
   "llama":
-    cmd: llama-server --port 8999 -m Llama-3.2-1B-Instruct-Q4_K_M.gguf
+    # multiline for readability
+    cmd: >
+      llama-server --port 8999
+      --model path/to/Qwen2.5-1.5B-Instruct-Q4_K_M.gguf
+
+    # environment variables to pass to the command
+    env:
+      - "CUDA_VISIBLE_DEVICES=0"
 
     # where to reach the server started by cmd, make sure the ports match
     proxy: http://127.0.0.1:8999
@@ -88,16 +101,9 @@ models:
     # default: 0 = never unload model
     ttl: 60
 
-  "qwen":
-    # environment variables to pass to the command
-    env:
-      - "CUDA_VISIBLE_DEVICES=0"
-
-    # multiline for readability
-    cmd: >
-      llama-server --port 8999
-      --model path/to/Qwen2.5-1.5B-Instruct-Q4_K_M.gguf
-    proxy: http://127.0.0.1:8999
+    # `useModelName` overrides the model name in the request
+    # and sends a specific name to the upstream server
+    useModelName: "qwen:qwq"
 
   # unlisted models do not show up in /v1/models or /upstream lists
   # but they can still be requested as normal
@@ -114,16 +120,58 @@ models:
       ghcr.io/ggerganov/llama.cpp:server
       --model '/models/Qwen2.5-Coder-0.5B-Instruct-Q4_K_M.gguf'
 
-# profiles make it easy to managing multi model (and gpu) configurations.
+# Groups provide advanced controls over model swapping behaviour. Using groups
+# some models can be kept loaded indefinitely, while others are swapped out.
 #
 # Tips:
-#  - each model must be listening on a unique address and port
-#  - the model name is in this format: "profile_name:model", like "coding:qwen"
-#  - the profile will load and unload all models in the profile at the same time
-profiles:
-  coding:
-    - "qwen"
-    - "llama"
+#
+#  - models must be defined above in the Models section
+#  - a model can only be a member of one group
+#  - group behaviour is controlled via the `swap`, `exclusive` and `persistent` fields
+#  - see issue #109 for details
+#
+# NOTE: the example below uses model names that are not defined above for demonstration purposes
+groups:
+  # group1 is the default behaviour of llama-swap where only one model is allowed
+  # to run a time across the whole llama-swap instance
+  "group1":
+    # swap controls the model swapping behaviour in within the group
+    # - true : only one model is allowed to run at a time
+    # - false: all models can run together, no swapping
+    swap: true
+
+    # exclusive controls how the group affects other groups
+    # - true: causes all other groups to unload their models when this group runs a model
+    # - false: does not affect other groups
+    exclusive: true
+
+    # members references the models defined above
+    members:
+      - "llama"
+      - "qwen-unlisted"
+
+  # models in this group are never unloaded
+  "group2":
+    swap: false
+    exclusive: false
+    members:
+      - "docker-llama"
+      # (not defined above, here for example)
+      - "modelA"
+      - "modelB"
+
+  "forever":
+    # setting persistent to true causes the group to never be affected by the swapping behaviour of
+    # other groups. It is a shortcut to keeping some models always loaded.
+    persistent: true
+
+    # set swap/exclusive to false to prevent swapping inside the group and effect on other groups
+    swap: false
+    exclusive: false
+    members:
+      - "forever-modelA"
+      - "forever-modelB"
+      - "forever-modelc"
 ```
 
 ### Use Case Examples
@@ -213,9 +261,15 @@ Of course, CLI access is also supported:
 # sends up to the last 10KB of logs
 curl http://host/logs'
 
-# streams logs
+# streams combined logs
 curl -Ns 'http://host/logs/stream'
 
+# just llama-swap's logs
+curl -Ns 'http://host/logs/stream/proxy'
+
+# just upstream's logs
+curl -Ns 'http://host/logs/stream/upstream'
+
 # stream and filter logs with linux pipes
 curl -Ns http://host/logs/stream | grep 'eval time'
 
@@ -257,8 +311,4 @@ WantedBy=multi-user.target
 
 ## Star History
 
-<picture>
-   <source media="(prefers-color-scheme: dark)" srcset="https://api.star-history.com/svg?repos=mostlygeek/llama-swap&type=Date&theme=dark" />
-   <source media="(prefers-color-scheme: light)" srcset="https://api.star-history.com/svg?repos=mostlygeek/llama-swap&type=Date" />
-   <img alt="Star History Chart" src="https://api.star-history.com/svg?repos=mostlygeek/llama-swap&type=Date" />
-</picture>
+[![Star History Chart](https://api.star-history.com/svg?repos=mostlygeek/llama-swap&type=Date)](https://www.star-history.com/#mostlygeek/llama-swap&Date)

config.example.yaml

@@ -1,9 +1,9 @@
 # Seconds to wait for llama.cpp to be available to serve requests
 # Default (and minimum): 15 seconds
-healthCheckTimeout: 15
+healthCheckTimeout: 90
 
-# Log HTTP requests helpful for troubleshoot, defaults to False
-logRequests: true
+# valid log levels: debug, info (default), warn, error
+logLevel: debug
 
 models:
   "llama":

docker/build-container.sh

@@ -38,6 +38,12 @@ else
         | jq -r --arg arch "$ARCH" '.[] | select(.metadata.container.tags[] | startswith("server-\($arch)")) | .metadata.container.tags[]' \
         | sort -r | head -n1 | awk -F '-' '{print $3}')
 
+    # Abort if LCPP_TAG is empty.
+    if [[ -z "$LCPP_TAG" ]]; then
+        echo "Abort: Could not find llama-server container for arch: $ARCH"
+        exit 1
+    fi
+
     CONTAINER_TAG="ghcr.io/mostlygeek/llama-swap:v${LS_VER}-${ARCH}-${LCPP_TAG}"
     CONTAINER_LATEST="ghcr.io/mostlygeek/llama-swap:${ARCH}"
     echo "Building ${CONTAINER_TAG} $LS_VER"

examples/aider-qwq-coder/README.md

@@ -0,0 +1,153 @@
+# aider, QwQ, Qwen-Coder 2.5 and llama-swap
+
+This guide show how to use aider and llama-swap to get a 100% local coding co-pilot setup. The focus is on the trickest part which is configuring aider, llama-swap and llama-server to work together.
+
+## Here's what you you need:
+
+- aider - [installation docs](https://aider.chat/docs/install.html)
+- llama-server - [download latest release](https://github.com/ggml-org/llama.cpp/releases)
+- llama-swap - [download latest release](https://github.com/mostlygeek/llama-swap/releases)
+- [QwQ 32B](https://huggingface.co/bartowski/Qwen_QwQ-32B-GGUF) and [Qwen Coder 2.5 32B](https://huggingface.co/bartowski/Qwen2.5-Coder-32B-Instruct-GGUF) models
+- 24GB VRAM video card
+
+## Running aider
+
+The goal is getting this command line to work:
+
+```sh
+aider --architect \
+    --no-show-model-warnings \
+    --model openai/QwQ \
+    --editor-model openai/qwen-coder-32B \
+    --model-settings-file aider.model.settings.yml \
+    --openai-api-key "sk-na" \
+    --openai-api-base "http://10.0.1.24:8080/v1" \
+```
+
+Set `--openai-api-base` to the IP and port where your llama-swap is running.
+
+## Create an aider model settings file
+
+```yaml
+# aider.model.settings.yml
+
+#
+# !!! important: model names must match llama-swap configuration names !!!
+#
+
+- name: "openai/QwQ"
+  edit_format: diff
+  extra_params:
+    max_tokens: 16384
+    top_p: 0.95
+    top_k: 40
+    presence_penalty: 0.1
+    repetition_penalty: 1
+    num_ctx: 16384
+  use_temperature: 0.6
+  reasoning_tag: think
+  weak_model_name: "openai/qwen-coder-32B"
+  editor_model_name: "openai/qwen-coder-32B"
+
+- name: "openai/qwen-coder-32B"
+  edit_format: diff
+  extra_params:
+    max_tokens: 16384
+    top_p: 0.8
+    top_k: 20
+    repetition_penalty: 1.05
+  use_temperature: 0.6
+  reasoning_tag: think
+  editor_edit_format: editor-diff
+  editor_model_name: "openai/qwen-coder-32B"
+```
+
+## llama-swap configuration
+
+```yaml
+# config.yaml
+
+# The parameters are tweaked to fit model+context into 24GB VRAM GPUs
+models:
+  "qwen-coder-32B":
+    proxy: "http://127.0.0.1:8999"
+    cmd: >
+      /path/to/llama-server
+      --host 127.0.0.1 --port 8999 --flash-attn --slots
+      --ctx-size 16000
+      --cache-type-k q8_0 --cache-type-v q8_0
+       -ngl 99
+      --model /path/to/Qwen2.5-Coder-32B-Instruct-Q4_K_M.gguf
+
+  "QwQ":
+    proxy: "http://127.0.0.1:9503"
+    cmd: >
+      /path/to/llama-server
+      --host 127.0.0.1 --port 9503 --flash-attn --metrics--slots
+      --cache-type-k q8_0 --cache-type-v q8_0
+      --ctx-size 32000
+      --samplers "top_k;top_p;min_p;temperature;dry;typ_p;xtc"
+      --temp 0.6 --repeat-penalty 1.1 --dry-multiplier 0.5
+      --min-p 0.01 --top-k 40 --top-p 0.95
+      -ngl 99
+      --model /mnt/nvme/models/bartowski/Qwen_QwQ-32B-Q4_K_M.gguf
+```
+
+## Advanced, Dual GPU Configuration
+
+If you have _dual 24GB GPUs_ you can use llama-swap profiles to avoid swapping between QwQ and Qwen Coder.
+
+In llama-swap's configuration file:
+
+1. add a `profiles` section with `aider` as the profile name
+2. using the `env` field to specify the GPU IDs for each model
+
+```yaml
+# config.yaml
+
+# Add a profile for aider
+profiles:
+  aider:
+    - qwen-coder-32B
+    - QwQ
+
+models:
+  "qwen-coder-32B":
+    # manually set the GPU to run on
+    env:
+      - "CUDA_VISIBLE_DEVICES=0"
+    proxy: "http://127.0.0.1:8999"
+    cmd: /path/to/llama-server ...
+
+  "QwQ":
+    # manually set the GPU to run on
+    env:
+      - "CUDA_VISIBLE_DEVICES=1"
+    proxy: "http://127.0.0.1:9503"
+    cmd: /path/to/llama-server ...
+```
+
+Append the profile tag, `aider:`, to the model names in the model settings file
+
+```yaml
+# aider.model.settings.yml
+- name: "openai/aider:QwQ"
+  weak_model_name: "openai/aider:qwen-coder-32B-aider"
+  editor_model_name: "openai/aider:qwen-coder-32B-aider"
+
+- name: "openai/aider:qwen-coder-32B"
+  editor_model_name: "openai/aider:qwen-coder-32B-aider"
+```
+
+Run aider with:
+
+```sh
+$ aider --architect \
+    --no-show-model-warnings \
+    --model openai/aider:QwQ \
+    --editor-model openai/aider:qwen-coder-32B \
+    --config aider.conf.yml \
+    --model-settings-file aider.model.settings.yml
+    --openai-api-key "sk-na" \
+    --openai-api-base "http://10.0.1.24:8080/v1"
+```

examples/aider-qwq-coder/aider.model.settings.dualgpu.yml

@@ -0,0 +1,28 @@
+# this makes use of llama-swap's profile feature to
+# keep the architect and editor models in VRAM on different GPUs
+
+- name: "openai/aider:QwQ"
+  edit_format: diff
+  extra_params:
+    max_tokens: 16384
+    top_p: 0.95
+    top_k: 40
+    presence_penalty: 0.1
+    repetition_penalty: 1
+    num_ctx: 16384
+  use_temperature: 0.6
+  reasoning_tag: think
+  weak_model_name: "openai/aider:qwen-coder-32B"
+  editor_model_name: "openai/aider:qwen-coder-32B"
+
+- name: "openai/aider:qwen-coder-32B"
+  edit_format: diff
+  extra_params:
+    max_tokens: 16384
+    top_p: 0.8
+    top_k: 20
+    repetition_penalty: 1.05
+  use_temperature: 0.6
+  reasoning_tag: think
+  editor_edit_format: editor-diff
+  editor_model_name: "openai/aider:qwen-coder-32B"

examples/aider-qwq-coder/aider.model.settings.yml

@@ -0,0 +1,26 @@
+- name: "openai/QwQ"
+  edit_format: diff
+  extra_params:
+    max_tokens: 16384
+    top_p: 0.95
+    top_k: 40
+    presence_penalty: 0.1
+    repetition_penalty: 1
+    num_ctx: 16384
+  use_temperature: 0.6
+  reasoning_tag: think
+  weak_model_name: "openai/qwen-coder-32B"
+  editor_model_name: "openai/qwen-coder-32B"
+
+- name: "openai/qwen-coder-32B"
+  edit_format: diff
+  extra_params:
+    max_tokens: 16384
+    top_p: 0.8
+    top_k: 20
+    repetition_penalty: 1.05
+  use_temperature: 0.6
+  reasoning_tag: think
+  editor_edit_format: editor-diff
+  editor_model_name: "openai/qwen-coder-32B"
+

examples/aider-qwq-coder/llama-swap.yaml

@@ -0,0 +1,49 @@
+healthCheckTimeout: 300
+logLevel: debug
+
+profiles:
+    aider:
+      - qwen-coder-32B
+      - QwQ
+
+models:
+  "qwen-coder-32B":
+    env:
+      - "CUDA_VISIBLE_DEVICES=0"
+    aliases:
+      - coder
+    proxy: "http://127.0.0.1:8999"
+
+    # set appropriate paths for your environment
+    cmd: >
+      /path/to/llama-server
+      --host 127.0.0.1 --port 8999 --flash-attn --slots
+      --ctx-size 16000
+      --ctx-size-draft 16000
+      --model /path/to/Qwen2.5-Coder-32B-Instruct-Q4_K_M.gguf
+      --model-draft /path/to/Qwen2.5-Coder-1.5B-Instruct-Q8_0.gguf
+      -ngl 99 -ngld 99
+      --draft-max 16 --draft-min 4 --draft-p-min 0.4
+      --cache-type-k q8_0 --cache-type-v q8_0
+  "QwQ":
+    env:
+      - "CUDA_VISIBLE_DEVICES=1"
+    proxy: "http://127.0.0.1:9503"
+
+    # set appropriate paths for your environment
+    cmd: >
+      /path/to/llama-server
+      --host 127.0.0.1 --port 9503
+      --flash-attn --metrics
+      --slots
+      --model /path/to/Qwen_QwQ-32B-Q4_K_M.gguf
+      --cache-type-k q8_0 --cache-type-v q8_0
+      --ctx-size 32000
+      --samplers "top_k;top_p;min_p;temperature;dry;typ_p;xtc"
+      --temp 0.6
+      --repeat-penalty 1.1
+      --dry-multiplier 0.5
+      --min-p 0.01
+      --top-k 40
+      --top-p 0.95
+      -ngl 99 -ngld 99

go.mod

@@ -3,7 +3,11 @@ module github.com/mostlygeek/llama-swap
 go 1.23.0
 
 require (
+	github.com/gin-gonic/gin v1.10.0
+	github.com/google/shlex v0.0.0-20191202100458-e7afc7fbc510
 	github.com/stretchr/testify v1.9.0
+	github.com/tidwall/gjson v1.18.0
+	github.com/tidwall/sjson v1.2.5
 	gopkg.in/yaml.v3 v3.0.1
 )
 
@@ -15,12 +19,10 @@ require (
 	github.com/davecgh/go-spew v1.1.1 // indirect
 	github.com/gabriel-vasile/mimetype v1.4.3 // indirect
 	github.com/gin-contrib/sse v0.1.0 // indirect
-	github.com/gin-gonic/gin v1.10.0 // indirect
 	github.com/go-playground/locales v0.14.1 // indirect
 	github.com/go-playground/universal-translator v0.18.1 // indirect
 	github.com/go-playground/validator/v10 v10.20.0 // indirect
 	github.com/goccy/go-json v0.10.2 // indirect
-	github.com/google/shlex v0.0.0-20191202100458-e7afc7fbc510 // indirect
 	github.com/json-iterator/go v1.1.12 // indirect
 	github.com/klauspost/cpuid/v2 v2.2.7 // indirect
 	github.com/leodido/go-urn v1.4.0 // indirect
@@ -29,16 +31,14 @@ require (
 	github.com/modern-go/reflect2 v1.0.2 // indirect
 	github.com/pelletier/go-toml/v2 v2.2.2 // indirect
 	github.com/pmezard/go-difflib v1.0.0 // indirect
-	github.com/tidwall/gjson v1.18.0 // indirect
 	github.com/tidwall/match v1.1.1 // indirect
 	github.com/tidwall/pretty v1.2.1 // indirect
-	github.com/tidwall/sjson v1.2.5 // indirect
 	github.com/twitchyliquid64/golang-asm v0.15.1 // indirect
 	github.com/ugorji/go/codec v1.2.12 // indirect
 	golang.org/x/arch v0.8.0 // indirect
-	golang.org/x/crypto v0.31.0 // indirect
-	golang.org/x/net v0.33.0 // indirect
-	golang.org/x/sys v0.28.0 // indirect
-	golang.org/x/text v0.21.0 // indirect
+	golang.org/x/crypto v0.36.0 // indirect
+	golang.org/x/net v0.38.0 // indirect
+	golang.org/x/sys v0.31.0 // indirect
+	golang.org/x/text v0.23.0 // indirect
 	google.golang.org/protobuf v1.34.1 // indirect
 )

go.sum

@@ -78,20 +78,30 @@ golang.org/x/crypto v0.23.0 h1:dIJU/v2J8Mdglj/8rJ6UUOM3Zc9zLZxVZwwxMooUSAI=
 golang.org/x/crypto v0.23.0/go.mod h1:CKFgDieR+mRhux2Lsu27y0fO304Db0wZe70UKqHu0v8=
 golang.org/x/crypto v0.31.0 h1:ihbySMvVjLAeSH1IbfcRTkD/iNscyz8rGzjF/E5hV6U=
 golang.org/x/crypto v0.31.0/go.mod h1:kDsLvtWBEx7MV9tJOj9bnXsPbxwJQ6csT/x4KIN4Ssk=
+golang.org/x/crypto v0.36.0 h1:AnAEvhDddvBdpY+uR+MyHmuZzzNqXSe/GvuDeob5L34=
+golang.org/x/crypto v0.36.0/go.mod h1:Y4J0ReaxCR1IMaabaSMugxJES1EpwhBHhv2bDHklZvc=
 golang.org/x/net v0.25.0 h1:d/OCCoBEUq33pjydKrGQhw7IlUPI2Oylr+8qLx49kac=
 golang.org/x/net v0.25.0/go.mod h1:JkAGAh7GEvH74S6FOH42FLoXpXbE/aqXSrIQjXgsiwM=
 golang.org/x/net v0.33.0 h1:74SYHlV8BIgHIFC/LrYkOGIwL19eTYXQ5wc6TBuO36I=
 golang.org/x/net v0.33.0/go.mod h1:HXLR5J+9DxmrqMwG9qjGCxZ+zKXxBru04zlTvWlWuN4=
+golang.org/x/net v0.37.0 h1:1zLorHbz+LYj7MQlSf1+2tPIIgibq2eL5xkrGk6f+2c=
+golang.org/x/net v0.37.0/go.mod h1:ivrbrMbzFq5J41QOQh0siUuly180yBYtLp+CKbEaFx8=
+golang.org/x/net v0.38.0 h1:vRMAPTMaeGqVhG5QyLJHqNDwecKTomGeqbnfZyKlBI8=
+golang.org/x/net v0.38.0/go.mod h1:ivrbrMbzFq5J41QOQh0siUuly180yBYtLp+CKbEaFx8=
 golang.org/x/sys v0.5.0/go.mod h1:oPkhp1MJrh7nUepCBck5+mAzfO9JrbApNNgaTdGDITg=
 golang.org/x/sys v0.6.0/go.mod h1:oPkhp1MJrh7nUepCBck5+mAzfO9JrbApNNgaTdGDITg=
 golang.org/x/sys v0.20.0 h1:Od9JTbYCk261bKm4M/mw7AklTlFYIa0bIp9BgSm1S8Y=
 golang.org/x/sys v0.20.0/go.mod h1:/VUhepiaJMQUp4+oa/7Zr1D23ma6VTLIYjOOTFZPUcA=
 golang.org/x/sys v0.28.0 h1:Fksou7UEQUWlKvIdsqzJmUmCX3cZuD2+P3XyyzwMhlA=
 golang.org/x/sys v0.28.0/go.mod h1:/VUhepiaJMQUp4+oa/7Zr1D23ma6VTLIYjOOTFZPUcA=
+golang.org/x/sys v0.31.0 h1:ioabZlmFYtWhL+TRYpcnNlLwhyxaM9kWTDEmfnprqik=
+golang.org/x/sys v0.31.0/go.mod h1:BJP2sWEmIv4KK5OTEluFJCKSidICx8ciO85XgH3Ak8k=
 golang.org/x/text v0.15.0 h1:h1V/4gjBv8v9cjcR6+AR5+/cIYK5N/WAgiv4xlsEtAk=
 golang.org/x/text v0.15.0/go.mod h1:18ZOQIKpY8NJVqYksKHtTdi31H5itFRjB5/qKTNYzSU=
 golang.org/x/text v0.21.0 h1:zyQAAkrwaneQ066sspRyJaG9VNi/YJ1NfzcGB3hZ/qo=
 golang.org/x/text v0.21.0/go.mod h1:4IBbMaMmOPCJ8SecivzSH54+73PCFmPWxNTLm+vZkEQ=
+golang.org/x/text v0.23.0 h1:D71I7dUrlY+VX0gQShAThNGHFxZ13dGLBHQLVl1mJlY=
+golang.org/x/text v0.23.0/go.mod h1:/BLNzu4aZCJ1+kcD0DNRotWKage4q2rGVAg4o22unh4=
 google.golang.org/protobuf v1.34.1 h1:9ddQBjfCyZPOHPUiPxpYESBLc+T8P3E+Vo4IbKZgFWg=
 google.golang.org/protobuf v1.34.1/go.mod h1:c6P6GXX6sHbq/GpV6MGZEdwhWPcYBgnhAHhKbcUYpos=
 gopkg.in/check.v1 v0.0.0-20161208181325-20d25e280405 h1:yhCVgyC4o1eVCa2tZl7eS0r+SDo693bJlVdllGtEeKM=

llama-swap.go

@@ -34,6 +34,10 @@ func main() {
 		os.Exit(1)
 	}
 
+	if len(config.Profiles) > 0 {
+		fmt.Println("WARNING: Profile functionality has been removed in favor of Groups. See the README for more information.")
+	}
+
 	if mode := os.Getenv("GIN_MODE"); mode != "" {
 		gin.SetMode(mode)
 	} else {

misc/simple-responder/simple-responder.go

@@ -67,6 +67,46 @@ func main() {
 		c.String(200, *responseMessage)
 	})
 
+	// issue #41
+	r.POST("/v1/audio/transcriptions", func(c *gin.Context) {
+		// Parse the multipart form
+		if err := c.Request.ParseMultipartForm(10 << 20); err != nil { // 10 MB max memory
+			c.JSON(http.StatusBadRequest, gin.H{"error": fmt.Sprintf("Error parsing multipart form: %s", err)})
+			return
+		}
+
+		// Get the model from the form values
+		model := c.Request.FormValue("model")
+
+		if model == "" {
+			c.JSON(http.StatusBadRequest, gin.H{"error": "Missing model parameter"})
+			return
+		}
+
+		// Get the file from the form
+		file, _, err := c.Request.FormFile("file")
+		if err != nil {
+			c.JSON(http.StatusBadRequest, gin.H{"error": fmt.Sprintf("Error getting file: %s", err)})
+			return
+		}
+		defer file.Close()
+
+		// Read the file content to get its size
+		fileBytes, err := io.ReadAll(file)
+		if err != nil {
+			c.JSON(http.StatusInternalServerError, gin.H{"error": fmt.Sprintf("Error reading file: %s", err)})
+			return
+		}
+
+		fileSize := len(fileBytes)
+
+		// Return a JSON response with the model and transcription text including file size
+		c.JSON(http.StatusOK, gin.H{
+			"text":  fmt.Sprintf("The length of the file is %d bytes", fileSize),
+			"model": model,
+		})
+	})
+
 	r.GET("/slow-respond", func(c *gin.Context) {
 		echo := c.Query("echo")
 		delay := c.Query("delay")

proxy/config.go

@@ -3,12 +3,15 @@ package proxy
 import (
 	"fmt"
 	"os"
+	"sort"
 	"strings"
 
 	"github.com/google/shlex"
 	"gopkg.in/yaml.v3"
 )
 
+const DEFAULT_GROUP_ID = "(default)"
+
 type ModelConfig struct {
 	Cmd           string   `yaml:"cmd"`
 	Proxy         string   `yaml:"proxy"`
@@ -17,17 +20,45 @@ type ModelConfig struct {
 	CheckEndpoint string   `yaml:"checkEndpoint"`
 	UnloadAfter   int      `yaml:"ttl"`
 	Unlisted      bool     `yaml:"unlisted"`
+	UseModelName  string   `yaml:"useModelName"`
 }
 
 func (m *ModelConfig) SanitizedCommand() ([]string, error) {
 	return SanitizeCommand(m.Cmd)
 }
 
+type GroupConfig struct {
+	Swap       bool     `yaml:"swap"`
+	Exclusive  bool     `yaml:"exclusive"`
+	Persistent bool     `yaml:"persistent"`
+	Members    []string `yaml:"members"`
+}
+
+// set default values for GroupConfig
+func (c *GroupConfig) UnmarshalYAML(unmarshal func(interface{}) error) error {
+	type rawGroupConfig GroupConfig
+	defaults := rawGroupConfig{
+		Swap:       true,
+		Exclusive:  true,
+		Persistent: false,
+		Members:    []string{},
+	}
+
+	if err := unmarshal(&defaults); err != nil {
+		return err
+	}
+
+	*c = GroupConfig(defaults)
+	return nil
+}
+
 type Config struct {
 	HealthCheckTimeout int                    `yaml:"healthCheckTimeout"`
 	LogRequests        bool                   `yaml:"logRequests"`
-	Models             map[string]ModelConfig `yaml:"models"`
+	LogLevel           string                 `yaml:"logLevel"`
+	Models             map[string]ModelConfig `yaml:"models"` /* key is model ID */
 	Profiles           map[string][]string    `yaml:"profiles"`
+	Groups             map[string]GroupConfig `yaml:"groups"` /* key is group ID */
 
 	// map aliases to actual model IDs
 	aliases map[string]string
@@ -51,16 +82,16 @@ func (c *Config) FindConfig(modelName string) (ModelConfig, string, bool) {
 	}
 }
 
-func LoadConfig(path string) (*Config, error) {
+func LoadConfig(path string) (Config, error) {
 	data, err := os.ReadFile(path)
 	if err != nil {
-		return nil, err
+		return Config{}, err
 	}
 
 	var config Config
 	err = yaml.Unmarshal(data, &config)
 	if err != nil {
-		return nil, err
+		return Config{}, err
 	}
 
 	if config.HealthCheckTimeout < 15 {
@@ -75,7 +106,72 @@ func LoadConfig(path string) (*Config, error) {
 		}
 	}
 
-	return &config, nil
+	config = AddDefaultGroupToConfig(config)
+	// check that members are all unique in the groups
+	memberUsage := make(map[string]string) // maps member to group it appears in
+	for groupID, groupConfig := range config.Groups {
+		prevSet := make(map[string]bool)
+		for _, member := range groupConfig.Members {
+			// Check for duplicates within this group
+			if _, found := prevSet[member]; found {
+				return Config{}, fmt.Errorf("duplicate model member %s found in group: %s", member, groupID)
+			}
+			prevSet[member] = true
+
+			// Check if member is used in another group
+			if existingGroup, exists := memberUsage[member]; exists {
+				return Config{}, fmt.Errorf("model member %s is used in multiple groups: %s and %s", member, existingGroup, groupID)
+			}
+			memberUsage[member] = groupID
+		}
+	}
+
+	return config, nil
+}
+
+// rewrites the yaml to include a default group with any orphaned models
+func AddDefaultGroupToConfig(config Config) Config {
+
+	if config.Groups == nil {
+		config.Groups = make(map[string]GroupConfig)
+	}
+
+	defaultGroup := GroupConfig{
+		Swap:      true,
+		Exclusive: true,
+		Members:   []string{},
+	}
+	// if groups is empty, create a default group and put
+	// all models into it
+	if len(config.Groups) == 0 {
+		for modelName := range config.Models {
+			defaultGroup.Members = append(defaultGroup.Members, modelName)
+		}
+	} else {
+		// iterate over existing group members and add non-grouped models into the default group
+		for modelName, _ := range config.Models {
+			foundModel := false
+		found:
+			// search for the model in existing groups
+			for _, groupConfig := range config.Groups {
+				for _, member := range groupConfig.Members {
+					if member == modelName {
+						foundModel = true
+						break found
+					}
+				}
+			}
+
+			if !foundModel {
+				defaultGroup.Members = append(defaultGroup.Members, modelName)
+			}
+		}
+	}
+
+	sort.Strings(defaultGroup.Members) // make consistent ordering for testing
+	config.Groups[DEFAULT_GROUP_ID] = defaultGroup
+
+	return config
 }
 
 func SanitizeCommand(cmdStr string) ([]string, error) {

proxy/config_test.go

@@ -35,11 +35,31 @@ models:
     aliases:
       - "m2"
     checkEndpoint: "/"
+  model3:
+    cmd: path/to/cmd --arg1 one
+    proxy: "http://localhost:8081"
+    aliases:
+      - "mthree"
+    checkEndpoint: "/"
+  model4:
+    cmd: path/to/cmd --arg1 one
+    checkEndpoint: "/"
+
 healthCheckTimeout: 15
 profiles:
   test:
     - model1
     - model2
+groups:
+  group1:
+    swap: true
+    exclusive: false
+    members: ["model2"]
+  forever:
+    exclusive: false
+    persistent: true
+    members:
+      - "model4"
 `
 
 	if err := os.WriteFile(tempFile, []byte(content), 0644); err != nil {
@@ -52,7 +72,7 @@ profiles:
 		t.Fatalf("Failed to load config: %v", err)
 	}
 
-	expected := &Config{
+	expected := Config{
 		Models: map[string]ModelConfig{
 			"model1": {
 				Cmd:           "path/to/cmd --arg1 one",
@@ -68,6 +88,17 @@ profiles:
 				Env:           nil,
 				CheckEndpoint: "/",
 			},
+			"model3": {
+				Cmd:           "path/to/cmd --arg1 one",
+				Proxy:         "http://localhost:8081",
+				Aliases:       []string{"mthree"},
+				Env:           nil,
+				CheckEndpoint: "/",
+			},
+			"model4": {
+				Cmd:           "path/to/cmd --arg1 one",
+				CheckEndpoint: "/",
+			},
 		},
 		HealthCheckTimeout: 15,
 		Profiles: map[string][]string{
@@ -77,6 +108,25 @@ profiles:
 			"m1":        "model1",
 			"model-one": "model1",
 			"m2":        "model2",
+			"mthree":    "model3",
+		},
+		Groups: map[string]GroupConfig{
+			DEFAULT_GROUP_ID: {
+				Swap:      true,
+				Exclusive: true,
+				Members:   []string{"model1", "model3"},
+			},
+			"group1": {
+				Swap:      true,
+				Exclusive: false,
+				Members:   []string{"model2"},
+			},
+			"forever": {
+				Swap:       true,
+				Exclusive:  false,
+				Persistent: true,
+				Members:    []string{"model4"},
+			},
 		},
 	}
 
@@ -87,6 +137,51 @@ profiles:
 	assert.Equal(t, "model1", realname)
 }
 
+func TestConfig_GroupMemberIsUnique(t *testing.T) {
+	// Create a temporary YAML file for testing
+	tempDir, err := os.MkdirTemp("", "test-config")
+	if err != nil {
+		t.Fatalf("Failed to create temporary directory: %v", err)
+	}
+	defer os.RemoveAll(tempDir)
+
+	tempFile := filepath.Join(tempDir, "config.yaml")
+	content := `
+models:
+  model1:
+    cmd: path/to/cmd --arg1 one
+    proxy: "http://localhost:8080"
+  model2:
+    cmd: path/to/cmd --arg1 one
+    proxy: "http://localhost:8081"
+    checkEndpoint: "/"
+  model3:
+    cmd: path/to/cmd --arg1 one
+    proxy: "http://localhost:8081"
+    checkEndpoint: "/"
+
+healthCheckTimeout: 15
+groups:
+  group1:
+    swap: true
+    exclusive: false
+    members: ["model2"]
+  group2:
+    swap: true
+    exclusive: false
+    members: ["model2"]
+`
+
+	if err := os.WriteFile(tempFile, []byte(content), 0644); err != nil {
+		t.Fatalf("Failed to write temporary file: %v", err)
+	}
+
+	// Load the config and verify
+	_, err = LoadConfig(tempFile)
+	assert.NotNil(t, err)
+
+}
+
 func TestConfig_ModelConfigSanitizedCommand(t *testing.T) {
 	config := &ModelConfig{
 		Cmd: `python model1.py \

proxy/helpers_test.go

@@ -14,6 +14,7 @@ import (
 var (
 	nextTestPort int = 12000
 	portMutex    sync.Mutex
+	testLogger   = NewLogMonitorWriter(os.Stdout)
 )
 
 // Check if the binary exists
@@ -26,6 +27,17 @@ func TestMain(m *testing.M) {
 
 	gin.SetMode(gin.TestMode)
 
+	switch os.Getenv("LOG_LEVEL") {
+	case "debug":
+		testLogger.SetLogLevel(LevelDebug)
+	case "warn":
+		testLogger.SetLogLevel(LevelWarn)
+	case "info":
+		testLogger.SetLogLevel(LevelInfo)
+	default:
+		testLogger.SetLogLevel(LevelWarn)
+	}
+
 	m.Run()
 }
 

proxy/html/logs.html

@@ -12,32 +12,65 @@
             flex-direction: column;
             font-family: "Courier New", Courier, monospace;
         }
-        #log-controls {
-            margin: 0.5em;
+        .log-container {
             display: flex;
-            align-items: center;
-            justify-content: space-between; /* Spaces out elements evenly */
-        }
-        #log-controls input {
-            flex: 1;
-        }
-        #log-controls input:focus {
-           outline: none; /* Ensures no outline is shown when the input is focused */
-        }
-        #log-stream {
             flex: 1;
+            gap: 0.5em;
             margin: 0.5em;
+            min-height: 0;
+        }
+        .log-column {
+            display: flex;
+            flex-direction: column;
+            flex: 1;
+            min-width: 0;
+            transition: flex 0.3s ease;
+        }
+        .log-column.minimized {
+            flex: 0.1;
+            max-width: 50px;
+            border: 1px solid #777;
+            color: green;
+        }
+        .log-controls {
+            display: grid;
+            grid-template-columns: 1fr auto;
+            gap: 0.5em;
+            margin-bottom: 0.5em;
+        }
+        .log-controls input {
+            width: 100%;
+            padding: 4px;
+        }
+        .log-controls input:focus {
+           outline: none;
+        }
+        .log-stream {
+            flex: 1;
             padding: 1em;
             background: #f4f4f4;
             overflow-y: auto;
-            white-space: pre-wrap; /* Ensures line wrapping */
-            word-wrap: break-word; /* Ensures long words wrap */
+            white-space: pre-wrap;
+            word-wrap: break-word;
+            min-height: 0;
         }
 
         .regex-error {
             background-color: #ff0000 !important;
         }
 
+        /* Make headers clickable and show pointer cursor */
+        h2 {
+            cursor: pointer;
+            user-select: none;
+            margin: 0 0 0.5em 0;
+            padding: 0.5em;
+        }
+
+        h2:hover {
+            background-color: rgba(0, 0, 0, 0.05);
+        }
+
         /* Dark mode styles */
         @media (prefers-color-scheme: dark) {
             body {
@@ -45,101 +78,182 @@
                 color: #fff;
             }
 
-            #log-stream {
+            .log-stream {
                 background: #444;
                 color: #fff;
             }
 
-            #log-controls input {
+            .log-controls input {
                 background: #555;
                 color: #fff;
                 border: 1px solid #777;
             }
 
-            #log-controls button {
+            .log-controls button {
                 background: #555;
                 color: #fff;
                 border: 1px solid #777;
             }
+
+            h2:hover {
+                background-color: rgba(255, 255, 255, 0.1);
+            }
+        }
+
+        /* Hide content when minimized */
+        .log-column.minimized .log-controls,
+        .log-column.minimized .log-stream {
+            display: none;
+        }
+
+        .log-column.minimized h2 {
+            writing-mode: vertical-rl;
+            text-orientation: mixed;
+            transform: rotate(180deg);
+            white-space: nowrap;
+            margin: auto;
         }
     </style>
 </head>
 <body>
-    <pre id="log-stream">Waiting for logs...</pre>
-    <div id="log-controls">
-        <input type="text" id="filter-input" placeholder="regex filter">
-        <button id="clear-button">clear</button>
+    <div class="log-container">
+        <div class="log-column">
+            <h2>Proxy Logs</h2>
+            <div class="log-controls">
+                <input type="text" id="proxy-filter-input" placeholder="proxy regex filter">
+                <button id="proxy-clear-button">clear</button>
+            </div>
+            <pre class="log-stream" id="proxy-log-stream">Waiting for proxy logs...</pre>
+        </div>
+        <div class="log-column minimized">
+            <h2>Upstream Logs</h2>
+            <div class="log-controls">
+                <input type="text" id="upstream-filter-input" placeholder="upstream regex filter">
+                <button id="upstream-clear-button">clear</button>
+            </div>
+            <pre class="log-stream" id="upstream-log-stream">Waiting for upstream logs...</pre>
+        </div>
     </div>
     <script>
-        const logStream = document.getElementById('log-stream');
-        const filterInput = document.getElementById('filter-input');
-        var logData = "";
-        let regexFilter = null;
+        class LogStream {
+            constructor(streamElement, filterInput, clearButton, endpoint) {
+                this.streamElement = streamElement;
+                this.filterInput = filterInput;
+                this.clearButton = clearButton;
+                this.endpoint = endpoint;
+                this.logData = "";
+                this.regexFilter = null;
+                this.eventSource = null;
 
-        function setupEventSource() {
-            if (typeof(EventSource) !== "undefined") {
-                const eventSource = new EventSource("/logs/streamSSE");
-
-                eventSource.onmessage = function(event) {
-                    logData += event.data;
-                    render()
-                };
-
-                eventSource.onerror = function(err) {
-                    logData = "EventSource failed: " + err.message;
-                };
-            } else {
-                logData = "SSE Not supported by this browser."
+                this.initialize();
             }
-        }
 
-        // poor-ai's react  ¯\_(ツ)_/¯
-        function render() {
-            if (regexFilter) {
-                const lines = logData.split('\n');
-                const filteredLines = lines.filter(line => {
-                    return regexFilter === null || regexFilter.test(line);
+            initialize() {
+                this.filterInput.addEventListener('input', () => this.updateFilter());
+                this.clearButton.addEventListener('click', () => {
+                    this.filterInput.value = "";
+                    this.regexFilter = null;
+                    this.render();
                 });
-
-                if (filteredLines.length > 0) {
-                    logStream.textContent = filteredLines.join('\n') + '\n';
-                } else {
-                    logStream.textContent = "";
-                }
-            } else {
-                logStream.textContent = logData;
+                this.setupEventSource();
             }
 
-            logStream.scrollTop = logStream.scrollHeight;
-        }
+            setupEventSource() {
+                if (typeof(EventSource) === "undefined") {
+                    this.logData = "SSE Not supported by this browser.";
+                    this.render();
+                    return;
+                }
+
+                const connect = () => {
+                    this.eventSource = new EventSource(this.endpoint);
+
+                    this.eventSource.onmessage = (event) => {
+                        this.logData += event.data;
+                        this.logData = this.logData.slice(-1024 * 100);
+                        this.render();
+                    };
+
+                    this.eventSource.onerror = (err) => {
+                        // Close the current connection
+                        this.eventSource.close();
+
+                        this.logData += "\nConnection lost. Retrying in 5 seconds...\n";
+                        this.render();
+
+                        // Attempt to reconnect after 5 seconds
+                        setTimeout(() => {
+                            this.logData += "Attempting to reconnect...\n";
+                            this.render();
+                            connect();
+                        }, 5000);
+                    };
+                };
+
+                // Initial connection
+                connect();
+            }
+
+            render() {
+                let content = this.logData;
+
+                if (this.regexFilter) {
+                    const lines = content.split('\n');
+                    const filteredLines = lines.filter(line => this.regexFilter.test(line));
+                    content = filteredLines.length > 0 ? filteredLines.join('\n') + '\n' : "";
+                }
+
+                this.streamElement.textContent = content;
+                this.streamElement.scrollTop = this.streamElement.scrollHeight;
+            }
+
+            updateFilter() {
+                const pattern = this.filterInput.value.trim();
+                this.filterInput.classList.remove('regex-error');
+
+                if (!pattern) {
+                    this.regexFilter = null;
+                    this.render();
+                    return;
+                }
 
-        function updateFilter() {
-            const pattern = filterInput.value.trim();
-            filterInput.classList.remove('regex-error');
-            if (pattern) {
                 try {
-                    regexFilter = new RegExp(pattern);
+                    this.regexFilter = new RegExp(pattern);
                 } catch (e) {
                     console.error("Invalid regex pattern:", e);
-                    regexFilter = null;
-                    filterInput.classList.add('regex-error');
-                    return
+                    this.regexFilter = null;
+                    this.filterInput.classList.add('regex-error');
+                    return;
                 }
-            } else {
-                regexFilter = null;
-            }
 
-            render();
+                this.render();
+            }
         }
 
-        filterInput.addEventListener('input', updateFilter);
-        document.getElementById('clear-button').addEventListener('click', () => {
-            filterInput.value = "";
-            regexFilter = null;
-            render();
+        // Initialize both log streams
+        document.addEventListener('DOMContentLoaded', () => {
+            new LogStream(
+                document.getElementById('proxy-log-stream'),
+                document.getElementById('proxy-filter-input'),
+                document.getElementById('proxy-clear-button'),
+                "/logs/streamSSE/proxy"
+            );
+
+            new LogStream(
+                document.getElementById('upstream-log-stream'),
+                document.getElementById('upstream-filter-input'),
+                document.getElementById('upstream-clear-button'),
+                "/logs/streamSSE/upstream"
+            );
+
+            // Initialize clickable headers
+            document.querySelectorAll('h2').forEach(header => {
+                header.addEventListener('click', () => {
+                    const column = header.closest('.log-column');
+                    column.classList.toggle('minimized');
+                });
+            });
         });
-        setupEventSource();
-        updateFilter();
     </script>
 </body>
 </html>

proxy/logMonitor.go

@@ -2,11 +2,21 @@ package proxy
 
 import (
 	"container/ring"
+	"fmt"
 	"io"
 	"os"
 	"sync"
 )
 
+type LogLevel int
+
+const (
+	LevelDebug LogLevel = iota
+	LevelInfo
+	LevelWarn
+	LevelError
+)
+
 type LogMonitor struct {
 	clients  map[chan []byte]bool
 	mu       sync.RWMutex
@@ -15,6 +25,10 @@ type LogMonitor struct {
 
 	// typically this can be os.Stdout
 	stdout io.Writer
+
+	// logging levels
+	level  LogLevel
+	prefix string
 }
 
 func NewLogMonitor() *LogMonitor {
@@ -26,6 +40,8 @@ func NewLogMonitorWriter(stdout io.Writer) *LogMonitor {
 		clients: make(map[chan []byte]bool),
 		buffer:  ring.New(10 * 1024), // keep 10KB of buffered logs
 		stdout:  stdout,
+		level:   LevelInfo,
+		prefix:  "",
 	}
 }
 
@@ -94,3 +110,77 @@ func (w *LogMonitor) broadcast(msg []byte) {
 		}
 	}
 }
+
+func (w *LogMonitor) SetPrefix(prefix string) {
+	w.mu.Lock()
+	defer w.mu.Unlock()
+	w.prefix = prefix
+}
+
+func (w *LogMonitor) SetLogLevel(level LogLevel) {
+	w.mu.Lock()
+	defer w.mu.Unlock()
+	w.level = level
+}
+
+func (w *LogMonitor) formatMessage(level string, msg string) []byte {
+	prefix := ""
+	if w.prefix != "" {
+		prefix = fmt.Sprintf("[%s] ", w.prefix)
+	}
+	return []byte(fmt.Sprintf("%s[%s] %s\n", prefix, level, msg))
+}
+
+func (w *LogMonitor) log(level LogLevel, msg string) {
+	if level < w.level {
+		return
+	}
+	w.Write(w.formatMessage(level.String(), msg))
+}
+
+func (w *LogMonitor) Debug(msg string) {
+	w.log(LevelDebug, msg)
+}
+
+func (w *LogMonitor) Info(msg string) {
+	w.log(LevelInfo, msg)
+}
+
+func (w *LogMonitor) Warn(msg string) {
+	w.log(LevelWarn, msg)
+}
+
+func (w *LogMonitor) Error(msg string) {
+	w.log(LevelError, msg)
+}
+
+func (w *LogMonitor) Debugf(format string, args ...interface{}) {
+	w.log(LevelDebug, fmt.Sprintf(format, args...))
+}
+
+func (w *LogMonitor) Infof(format string, args ...interface{}) {
+	w.log(LevelInfo, fmt.Sprintf(format, args...))
+}
+
+func (w *LogMonitor) Warnf(format string, args ...interface{}) {
+	w.log(LevelWarn, fmt.Sprintf(format, args...))
+}
+
+func (w *LogMonitor) Errorf(format string, args ...interface{}) {
+	w.log(LevelError, fmt.Sprintf(format, args...))
+}
+
+func (l LogLevel) String() string {
+	switch l {
+	case LevelDebug:
+		return "DEBUG"
+	case LevelInfo:
+		return "INFO"
+	case LevelWarn:
+		return "WARN"
+	case LevelError:
+		return "ERROR"
+	default:
+		return "UNKNOWN"
+	}
+}

proxy/process.go

@@ -30,11 +30,18 @@ const (
 )
 
 type Process struct {
-	ID                 string
-	config             ModelConfig
-	cmd                *exec.Cmd
-	logMonitor         *LogMonitor
-	healthCheckTimeout int
+	ID     string
+	config ModelConfig
+	cmd    *exec.Cmd
+
+	// for p.cmd.Wait() select { ... }
+	cmdWaitChan chan error
+
+	processLogger *LogMonitor
+	proxyLogger   *LogMonitor
+
+	healthCheckTimeout      int
+	healthCheckLoopInterval time.Duration
 
 	lastRequestHandled time.Time
 
@@ -51,54 +58,70 @@ type Process struct {
 	shutdownCancel context.CancelFunc
 }
 
-func NewProcess(ID string, healthCheckTimeout int, config ModelConfig, logMonitor *LogMonitor) *Process {
+func NewProcess(ID string, healthCheckTimeout int, config ModelConfig, processLogger *LogMonitor, proxyLogger *LogMonitor) *Process {
 	ctx, cancel := context.WithCancel(context.Background())
 	return &Process{
-		ID:                 ID,
-		config:             config,
-		cmd:                nil,
-		logMonitor:         logMonitor,
-		healthCheckTimeout: healthCheckTimeout,
-		state:              StateStopped,
-		shutdownCtx:        ctx,
-		shutdownCancel:     cancel,
+		ID:                      ID,
+		config:                  config,
+		cmd:                     nil,
+		cmdWaitChan:             make(chan error, 1),
+		processLogger:           processLogger,
+		proxyLogger:             proxyLogger,
+		healthCheckTimeout:      healthCheckTimeout,
+		healthCheckLoopInterval: 5 * time.Second, /* default, can not be set by user - used for testing */
+		state:                   StateStopped,
+		shutdownCtx:             ctx,
+		shutdownCancel:          cancel,
 	}
 }
 
-func (p *Process) setState(newState ProcessState) error {
-	// enforce valid state transitions
-	invalidTransition := false
-	if p.state == StateStopped {
-		// stopped -> starting
-		if newState != StateStarting {
-			invalidTransition = true
-		}
-	} else if p.state == StateStarting {
-		// starting -> ready | failed | stopping
-		if newState != StateReady && newState != StateFailed && newState != StateStopping {
-			invalidTransition = true
-		}
-	} else if p.state == StateReady {
-		// ready -> stopping
-		if newState != StateStopping {
-			invalidTransition = true
-		}
-	} else if p.state == StateStopping {
-		// stopping -> stopped | shutdown
-		if newState != StateStopped && newState != StateShutdown {
-			invalidTransition = true
-		}
-	} else if p.state == StateFailed || p.state == StateShutdown {
-		invalidTransition = true
+// LogMonitor returns the log monitor associated with the process.
+func (p *Process) LogMonitor() *LogMonitor {
+	return p.processLogger
+}
+
+// custom error types for swapping state
+var (
+	ErrExpectedStateMismatch  = errors.New("expected state mismatch")
+	ErrInvalidStateTransition = errors.New("invalid state transition")
+)
+
+// swapState performs a compare and swap of the state atomically. It returns the current state
+// and an error if the swap failed.
+func (p *Process) swapState(expectedState, newState ProcessState) (ProcessState, error) {
+	p.stateMutex.Lock()
+	defer p.stateMutex.Unlock()
+
+	if p.state != expectedState {
+		p.proxyLogger.Warnf("<%s> swapState() Unexpected current state %s, expected %s", p.ID, p.state, expectedState)
+		return p.state, ErrExpectedStateMismatch
 	}
 
-	if invalidTransition {
-		//panic(fmt.Sprintf("Invalid state transition from %s to %s", p.state, newState))
-		return fmt.Errorf("invalid state transition from %s to %s", p.state, newState)
+	if !isValidTransition(p.state, newState) {
+		p.proxyLogger.Warnf("<%s> swapState() Invalid state transition from %s to %s", p.ID, p.state, newState)
+		return p.state, ErrInvalidStateTransition
 	}
 
 	p.state = newState
-	return nil
+	p.proxyLogger.Debugf("<%s> swapState() State transitioned from %s to %s", p.ID, expectedState, newState)
+	return p.state, nil
+}
+
+// Helper function to encapsulate transition rules
+func isValidTransition(from, to ProcessState) bool {
+	switch from {
+	case StateStopped:
+		return to == StateStarting
+	case StateStarting:
+		return to == StateReady || to == StateFailed || to == StateStopping
+	case StateReady:
+		return to == StateStopping
+	case StateStopping:
+		return to == StateStopped || to == StateShutdown
+	case StateFailed, StateShutdown:
+		return false // No transitions allowed from these states
+	}
+	return false
 }
 
 func (p *Process) CurrentState() ProcessState {
@@ -116,50 +139,58 @@ func (p *Process) start() error {
 		return fmt.Errorf("can not start(), upstream proxy missing")
 	}
 
-	// wait for the other start() to complete
-	curState := p.CurrentState()
-
-	if curState == StateReady {
-		return nil
-	}
-
-	if curState == StateStarting {
-		p.waitStarting.Wait()
-
-		if state := p.CurrentState(); state != StateReady {
-			return fmt.Errorf("start() failed current state: %v", state)
-		}
-
-		return nil
-	}
-
-	p.stateMutex.Lock()
-	defer p.stateMutex.Unlock()
-
-	if err := p.setState(StateStarting); err != nil {
-		return err
-	}
-
-	p.waitStarting.Add(1)
-	defer p.waitStarting.Done()
-
 	args, err := p.config.SanitizedCommand()
 	if err != nil {
 		return fmt.Errorf("unable to get sanitized command: %v", err)
 	}
 
+	if curState, err := p.swapState(StateStopped, StateStarting); err != nil {
+		if err == ErrExpectedStateMismatch {
+			// already starting, just wait for it to complete and expect
+			// it to be be in the Ready start after. If not, return an error
+			if curState == StateStarting {
+				p.waitStarting.Wait()
+				if state := p.CurrentState(); state == StateReady {
+					return nil
+				} else {
+					return fmt.Errorf("process was already starting but wound up in state %v", state)
+				}
+			} else {
+				return fmt.Errorf("processes was in state %v when start() was called", curState)
+			}
+		} else {
+			return fmt.Errorf("failed to set Process state to starting: current state: %v, error: %v", curState, err)
+		}
+	}
+
+	p.waitStarting.Add(1)
+	defer p.waitStarting.Done()
+
 	p.cmd = exec.Command(args[0], args[1:]...)
-	p.cmd.Stdout = p.logMonitor
-	p.cmd.Stderr = p.logMonitor
+	p.cmd.Stdout = p.processLogger
+	p.cmd.Stderr = p.processLogger
 	p.cmd.Env = p.config.Env
 
 	err = p.cmd.Start()
 
+	// Set process state to failed
 	if err != nil {
-		p.setState(StateFailed)
+		if curState, swapErr := p.swapState(StateStarting, StateFailed); swapErr != nil {
+			return fmt.Errorf(
+				"failed to start command and state swap failed. command error: %v, current state: %v, state swap error: %v",
+				err, curState, swapErr,
+			)
+		}
 		return fmt.Errorf("start() failed: %v", err)
 	}
 
+	// Capture the exit error for later signaling
+	go func() {
+		exitErr := p.cmd.Wait()
+		p.proxyLogger.Debugf("<%s> cmd.Wait() returned error: %v", p.ID, exitErr)
+		p.cmdWaitChan <- exitErr
+	}()
+
 	// One of three things can happen at this stage:
 	// 1. The command exits unexpectedly
 	// 2. The health check fails
@@ -191,31 +222,51 @@ func (p *Process) start() error {
 		)
 		defer cancelHealthCheck()
 
-		// Health check loop
 	loop:
+		// Ready Check loop
 		for {
 			select {
 			case <-checkDeadline.Done():
-				p.setState(StateFailed)
-				return fmt.Errorf("health check failed after %vs", maxDuration.Seconds())
+				if curState, err := p.swapState(StateStarting, StateFailed); err != nil {
+					return fmt.Errorf("health check timed out after %vs AND state swap failed: %v, current state: %v", maxDuration.Seconds(), err, curState)
+				} else {
+					return fmt.Errorf("health check timed out after %vs", maxDuration.Seconds())
+				}
 			case <-p.shutdownCtx.Done():
 				return errors.New("health check interrupted due to shutdown")
+			case exitErr := <-p.cmdWaitChan:
+				if exitErr != nil {
+					p.proxyLogger.Warnf("<%s> upstream command exited prematurely with error: %v", p.ID, exitErr)
+					if curState, err := p.swapState(StateStarting, StateFailed); err != nil {
+						return fmt.Errorf("upstream command exited unexpectedly: %s AND state swap failed: %v, current state: %v", exitErr.Error(), err, curState)
+					} else {
+						return fmt.Errorf("upstream command exited unexpectedly: %s", exitErr.Error())
+					}
+				} else {
+					p.proxyLogger.Warnf("<%s> upstream command exited prematurely but successfully", p.ID)
+					if curState, err := p.swapState(StateStarting, StateFailed); err != nil {
+						return fmt.Errorf("upstream command exited prematurely but successfully AND state swap failed: %v, current state: %v", err, curState)
+					} else {
+						return fmt.Errorf("upstream command exited prematurely but successfully")
+					}
+				}
 			default:
 				if err := p.checkHealthEndpoint(healthURL); err == nil {
+					p.proxyLogger.Infof("<%s> Health check passed on %s", p.ID, healthURL)
 					cancelHealthCheck()
 					break loop
 				} else {
 					if strings.Contains(err.Error(), "connection refused") {
 						endTime, _ := checkDeadline.Deadline()
 						ttl := time.Until(endTime)
-						fmt.Fprintf(p.logMonitor, "!!! Connection refused on %s, ttl %.0fs\n", healthURL, ttl.Seconds())
+						p.proxyLogger.Infof("<%s> Connection refused on %s, giving up in %.0fs", p.ID, healthURL, ttl.Seconds())
 					} else {
-						fmt.Fprintf(p.logMonitor, "!!! Health check error: %v\n", err)
+						p.proxyLogger.Infof("<%s> Health check error on %s, %v", p.ID, healthURL, err)
 					}
 				}
 			}
 
-			<-time.After(5 * time.Second)
+			<-time.After(p.healthCheckLoopInterval)
 		}
 	}
 
@@ -226,7 +277,7 @@ func (p *Process) start() error {
 			maxDuration := time.Duration(p.config.UnloadAfter) * time.Second
 
 			for range time.Tick(time.Second) {
-				if p.state != StateReady {
+				if p.CurrentState() != StateReady {
 					return
 				}
 
@@ -234,7 +285,7 @@ func (p *Process) start() error {
 				p.inFlightRequests.Wait()
 
 				if time.Since(p.lastRequestHandled) > maxDuration {
-					fmt.Fprintf(p.logMonitor, "!!! Unloading model %s, TTL of %ds reached.\n", p.ID, p.config.UnloadAfter)
+					p.proxyLogger.Infof("<%s> Unloading model, TTL of %ds reached", p.ID, p.config.UnloadAfter)
 					p.Stop()
 					return
 				}
@@ -242,26 +293,33 @@ func (p *Process) start() error {
 		}()
 	}
 
-	return p.setState(StateReady)
+	if curState, err := p.swapState(StateStarting, StateReady); err != nil {
+		return fmt.Errorf("failed to set Process state to ready: current state: %v, error: %v", curState, err)
+	} else {
+		return nil
+	}
 }
 
 func (p *Process) Stop() {
+	if !isValidTransition(p.CurrentState(), StateStopping) {
+		return
+	}
+
 	// wait for any inflight requests before proceeding
 	p.inFlightRequests.Wait()
-	p.stateMutex.Lock()
-	defer p.stateMutex.Unlock()
+	p.proxyLogger.Debugf("<%s> Stopping process", p.ID)
 
 	// calling Stop() when state is invalid is a no-op
-	if err := p.setState(StateStopping); err != nil {
-		fmt.Fprintf(p.logMonitor, "!!! Info - Stop() err: %v\n", err)
+	if curState, err := p.swapState(StateReady, StateStopping); err != nil {
+		p.proxyLogger.Infof("<%s> Stop() Ready -> StateStopping err: %v, current state: %v", p.ID, err, curState)
 		return
 	}
 
 	// stop the process with a graceful exit timeout
 	p.stopCommand(5 * time.Second)
 
-	if err := p.setState(StateStopped); err != nil {
-		panic(fmt.Sprintf("Stop() failed to set state to stopped: %v", err))
+	if curState, err := p.swapState(StateStopping, StateStopped); err != nil {
+		p.proxyLogger.Infof("<%s> Stop() StateStopping -> StateStopped err: %v, current state: %v", p.ID, err, curState)
 	}
 }
 
@@ -269,65 +327,59 @@ func (p *Process) Stop() {
 // of time for any inflight requests to complete before shutting down. If the Process
 // is in the state of starting, it will cancel it and shut it down
 func (p *Process) Shutdown() {
-	// cancel anything that can be interrupted by a shutdown (ie: healthcheck)
 	p.shutdownCancel()
-
-	p.stateMutex.Lock()
-	defer p.stateMutex.Unlock()
-	p.setState(StateStopping)
-
-	// 5 seconds to stop the process
 	p.stopCommand(5 * time.Second)
-	if err := p.setState(StateShutdown); err != nil {
-		fmt.Printf("!!! Shutdown() failed to set state to shutdown: %v", err)
-	}
-	p.setState(StateShutdown)
+	p.state = StateShutdown
 }
 
 // stopCommand will send a SIGTERM to the process and wait for it to exit.
 // If it does not exit within 5 seconds, it will send a SIGKILL.
 func (p *Process) stopCommand(sigtermTTL time.Duration) {
+	stopStartTime := time.Now()
+	defer func() {
+		p.proxyLogger.Debugf("<%s> stopCommand took %v", p.ID, time.Since(stopStartTime))
+	}()
+
 	sigtermTimeout, cancelTimeout := context.WithTimeout(context.Background(), sigtermTTL)
 	defer cancelTimeout()
 
-	sigtermNormal := make(chan error, 1)
-	go func() {
-		sigtermNormal <- p.cmd.Wait()
-	}()
-
 	if p.cmd == nil || p.cmd.Process == nil {
-		fmt.Fprintf(p.logMonitor, "!!! process [%s] cmd or cmd.Process is nil", p.ID)
+		p.proxyLogger.Warnf("<%s> cmd or cmd.Process is nil", p.ID)
 		return
 	}
 
-	p.cmd.Process.Signal(syscall.SIGTERM)
+	if err := p.terminateProcess(); err != nil {
+		p.proxyLogger.Infof("<%s> Failed to gracefully terminate process: %v", p.ID, err)
+	}
 
 	select {
 	case <-sigtermTimeout.Done():
-		fmt.Fprintf(p.logMonitor, "!!! process [%s] timed out waiting to stop, sending KILL signal\n", p.ID)
+		p.proxyLogger.Infof("<%s> Process timed out waiting to stop, sending KILL signal", p.ID)
 		p.cmd.Process.Kill()
-	case err := <-sigtermNormal:
+	case err := <-p.cmdWaitChan:
+		// Note: in start(), p.cmdWaitChan also has a select { ... }. That should be OK
+		// because if we make it here then the cmd has been successfully running and made it
+		// through the health check. There is a possibility that ithe cmd crashed after the health check
+		// succeeded but that's not a case llama-swap is handling for now.
 		if err != nil {
 			if errno, ok := err.(syscall.Errno); ok {
-				fmt.Fprintf(p.logMonitor, "!!! process [%s] errno >> %v\n", p.ID, errno)
+				p.proxyLogger.Errorf("<%s> errno >> %v", p.ID, errno)
 			} else if exitError, ok := err.(*exec.ExitError); ok {
 				if strings.Contains(exitError.String(), "signal: terminated") {
-					fmt.Fprintf(p.logMonitor, "!!! process [%s] stopped OK\n", p.ID)
+					p.proxyLogger.Infof("<%s> Process stopped OK", p.ID)
 				} else if strings.Contains(exitError.String(), "signal: interrupt") {
-					fmt.Fprintf(p.logMonitor, "!!! process [%s] interrupted OK\n", p.ID)
+					p.proxyLogger.Infof("<%s> Process interrupted OK", p.ID)
 				} else {
-					fmt.Fprintf(p.logMonitor, "!!! process [%s] ExitError >> %v, exit code: %d\n", p.ID, exitError, exitError.ExitCode())
+					p.proxyLogger.Warnf("<%s> ExitError >> %v, exit code: %d", p.ID, exitError, exitError.ExitCode())
 				}
-
 			} else {
-				fmt.Fprintf(p.logMonitor, "!!! process [%s] exited >> %v\n", p.ID, err)
+				p.proxyLogger.Errorf("<%s> Process exited >> %v", p.ID, err)
 			}
 		}
 	}
 }
 
 func (p *Process) checkHealthEndpoint(healthURL string) error {
-
 	client := &http.Client{
 		Timeout: 500 * time.Millisecond,
 	}
@@ -352,6 +404,8 @@ func (p *Process) checkHealthEndpoint(healthURL string) error {
 }
 
 func (p *Process) ProxyRequest(w http.ResponseWriter, r *http.Request) {
+	requestBeginTime := time.Now()
+	var startDuration time.Duration
 
 	// prevent new requests from being made while stopping or irrecoverable
 	currentState := p.CurrentState()
@@ -368,11 +422,13 @@ func (p *Process) ProxyRequest(w http.ResponseWriter, r *http.Request) {
 
 	// start the process on demand
 	if p.CurrentState() != StateReady {
+		beginStartTime := time.Now()
 		if err := p.start(); err != nil {
 			errstr := fmt.Sprintf("unable to start process: %s", err)
 			http.Error(w, errstr, http.StatusBadGateway)
 			return
 		}
+		startDuration = time.Since(beginStartTime)
 	}
 
 	proxyTo := p.config.Proxy
@@ -416,4 +472,8 @@ func (p *Process) ProxyRequest(w http.ResponseWriter, r *http.Request) {
 			return
 		}
 	}
+
+	totalTime := time.Since(requestBeginTime)
+	p.proxyLogger.Debugf("<%s> request %s - start: %v, total: %v",
+		p.ID, r.RequestURI, startDuration, totalTime)
 }

proxy/process_stop.go

@@ -0,0 +1,9 @@
+//go:build !windows
+
+package proxy
+
+import "syscall"
+
+func (p *Process) terminateProcess() error {
+	return p.cmd.Process.Signal(syscall.SIGTERM)
+}

proxy/process_stop_windows.go

@@ -0,0 +1,14 @@
+//go:build windows
+
+package proxy
+
+import (
+	"fmt"
+	"os/exec"
+)
+
+func (p *Process) terminateProcess() error {
+	pid := fmt.Sprintf("%d", p.cmd.Process.Pid)
+	cmd := exec.Command("taskkill", "/f", "/t", "/pid", pid)
+	return cmd.Run()
+}

proxy/process_test.go

@@ -2,7 +2,6 @@ package proxy
 
 import (
 	"fmt"
-	"io"
 	"net/http"
 	"net/http/httptest"
 	"os"
@@ -13,13 +12,26 @@ import (
 	"github.com/stretchr/testify/assert"
 )
 
+var (
+	debugLogger = NewLogMonitorWriter(os.Stdout)
+)
+
+func init() {
+	// flip to help with debugging tests
+	if false {
+		debugLogger.SetLogLevel(LevelDebug)
+	} else {
+		debugLogger.SetLogLevel(LevelError)
+	}
+}
+
 func TestProcess_AutomaticallyStartsUpstream(t *testing.T) {
-	logMonitor := NewLogMonitorWriter(io.Discard)
+
 	expectedMessage := "testing91931"
 	config := getTestSimpleResponderConfig(expectedMessage)
 
 	// Create a process
-	process := NewProcess("test-process", 5, config, logMonitor)
+	process := NewProcess("test-process", 5, config, debugLogger, debugLogger)
 	defer process.Stop()
 
 	req := httptest.NewRequest("GET", "/test", nil)
@@ -52,11 +64,10 @@ func TestProcess_AutomaticallyStartsUpstream(t *testing.T) {
 // are all handled successfully, even though they all may ask for the process to .start()
 func TestProcess_WaitOnMultipleStarts(t *testing.T) {
 
-	logMonitor := NewLogMonitorWriter(io.Discard)
 	expectedMessage := "testing91931"
 	config := getTestSimpleResponderConfig(expectedMessage)
 
-	process := NewProcess("test-process", 5, config, logMonitor)
+	process := NewProcess("test-process", 5, config, debugLogger, debugLogger)
 	defer process.Stop()
 
 	var wg sync.WaitGroup
@@ -84,7 +95,7 @@ func TestProcess_BrokenModelConfig(t *testing.T) {
 		CheckEndpoint: "/health",
 	}
 
-	process := NewProcess("broken", 1, config, NewLogMonitor())
+	process := NewProcess("broken", 1, config, debugLogger, debugLogger)
 
 	req := httptest.NewRequest("GET", "/", nil)
 	w := httptest.NewRecorder()
@@ -109,7 +120,7 @@ func TestProcess_UnloadAfterTTL(t *testing.T) {
 	config.UnloadAfter = 3 // seconds
 	assert.Equal(t, 3, config.UnloadAfter)
 
-	process := NewProcess("ttl_test", 2, config, NewLogMonitorWriter(io.Discard))
+	process := NewProcess("ttl_test", 2, config, debugLogger, debugLogger)
 	defer process.Stop()
 
 	// this should take 4 seconds
@@ -151,7 +162,7 @@ func TestProcess_LowTTLValue(t *testing.T) {
 	config.UnloadAfter = 1 // second
 	assert.Equal(t, 1, config.UnloadAfter)
 
-	process := NewProcess("ttl", 2, config, NewLogMonitorWriter(os.Stdout))
+	process := NewProcess("ttl", 2, config, debugLogger, debugLogger)
 	defer process.Stop()
 
 	for i := 0; i < 100; i++ {
@@ -169,6 +180,8 @@ func TestProcess_LowTTLValue(t *testing.T) {
 }
 
 // issue #19
+// This test makes sure using Process.Stop() does not affect pending HTTP
+// requests. All HTTP requests in this test should complete successfully.
 func TestProcess_HTTPRequestsHaveTimeToFinish(t *testing.T) {
 	if testing.Short() {
 		t.Skip("skipping slow test")
@@ -176,7 +189,7 @@ func TestProcess_HTTPRequestsHaveTimeToFinish(t *testing.T) {
 
 	expectedMessage := "12345"
 	config := getTestSimpleResponderConfig(expectedMessage)
-	process := NewProcess("t", 10, config, NewLogMonitorWriter(os.Stdout))
+	process := NewProcess("t", 10, config, debugLogger, debugLogger)
 	defer process.Stop()
 
 	results := map[string]string{
@@ -192,8 +205,9 @@ func TestProcess_HTTPRequestsHaveTimeToFinish(t *testing.T) {
 		wg.Add(1)
 		go func(key string) {
 			defer wg.Done()
-			// send a request that should take 5 * 200ms (1 second) to complete
-			req := httptest.NewRequest("GET", fmt.Sprintf("/slow-respond?echo=%s&delay=200ms", key), nil)
+			// send a request where simple-responder is will wait 300ms before responding
+			// this will simulate an in-progress request.
+			req := httptest.NewRequest("GET", fmt.Sprintf("/slow-respond?echo=%s&delay=300ms", key), nil)
 			w := httptest.NewRecorder()
 
 			process.ProxyRequest(w, req)
@@ -209,9 +223,9 @@ func TestProcess_HTTPRequestsHaveTimeToFinish(t *testing.T) {
 		}(key)
 	}
 
-	// stop the requests in the middle
+	// Stop the process while requests are still being processed
 	go func() {
-		<-time.After(500 * time.Millisecond)
+		<-time.After(150 * time.Millisecond)
 		process.Stop()
 	}()
 
@@ -222,39 +236,40 @@ func TestProcess_HTTPRequestsHaveTimeToFinish(t *testing.T) {
 	}
 }
 
-func TestSetState(t *testing.T) {
+func TestProcess_SwapState(t *testing.T) {
 	tests := []struct {
 		name           string
 		currentState   ProcessState
+		expectedState  ProcessState
 		newState       ProcessState
 		expectedError  error
 		expectedResult ProcessState
 	}{
-		{"Stopped to Starting", StateStopped, StateStarting, nil, StateStarting},
-		{"Starting to Ready", StateStarting, StateReady, nil, StateReady},
-		{"Starting to Failed", StateStarting, StateFailed, nil, StateFailed},
-		{"Starting to Stopping", StateStarting, StateStopping, nil, StateStopping},
-		{"Ready to Stopping", StateReady, StateStopping, nil, StateStopping},
-		{"Stopping to Stopped", StateStopping, StateStopped, nil, StateStopped},
-		{"Stopping to Shutdown", StateStopping, StateShutdown, nil, StateShutdown},
-		{"Stopped to Ready", StateStopped, StateReady, fmt.Errorf("invalid state transition from stopped to ready"), StateStopped},
-		{"Starting to Stopped", StateStarting, StateStopped, fmt.Errorf("invalid state transition from starting to stopped"), StateStarting},
-		{"Ready to Starting", StateReady, StateStarting, fmt.Errorf("invalid state transition from ready to starting"), StateReady},
-		{"Ready to Failed", StateReady, StateFailed, fmt.Errorf("invalid state transition from ready to failed"), StateReady},
-		{"Stopping to Ready", StateStopping, StateReady, fmt.Errorf("invalid state transition from stopping to ready"), StateStopping},
-		{"Failed to Stopped", StateFailed, StateStopped, fmt.Errorf("invalid state transition from failed to stopped"), StateFailed},
-		{"Failed to Starting", StateFailed, StateStarting, fmt.Errorf("invalid state transition from failed to starting"), StateFailed},
-		{"Shutdown to Stopped", StateShutdown, StateStopped, fmt.Errorf("invalid state transition from shutdown to stopped"), StateShutdown},
-		{"Shutdown to Starting", StateShutdown, StateStarting, fmt.Errorf("invalid state transition from shutdown to starting"), StateShutdown},
+		{"Stopped to Starting", StateStopped, StateStopped, StateStarting, nil, StateStarting},
+		{"Starting to Ready", StateStarting, StateStarting, StateReady, nil, StateReady},
+		{"Starting to Failed", StateStarting, StateStarting, StateFailed, nil, StateFailed},
+		{"Starting to Stopping", StateStarting, StateStarting, StateStopping, nil, StateStopping},
+		{"Ready to Stopping", StateReady, StateReady, StateStopping, nil, StateStopping},
+		{"Stopping to Stopped", StateStopping, StateStopping, StateStopped, nil, StateStopped},
+		{"Stopping to Shutdown", StateStopping, StateStopping, StateShutdown, nil, StateShutdown},
+		{"Stopped to Ready", StateStopped, StateStopped, StateReady, ErrInvalidStateTransition, StateStopped},
+		{"Starting to Stopped", StateStarting, StateStarting, StateStopped, ErrInvalidStateTransition, StateStarting},
+		{"Ready to Starting", StateReady, StateReady, StateStarting, ErrInvalidStateTransition, StateReady},
+		{"Ready to Failed", StateReady, StateReady, StateFailed, ErrInvalidStateTransition, StateReady},
+		{"Stopping to Ready", StateStopping, StateStopping, StateReady, ErrInvalidStateTransition, StateStopping},
+		{"Failed to Stopped", StateFailed, StateFailed, StateStopped, ErrInvalidStateTransition, StateFailed},
+		{"Failed to Starting", StateFailed, StateFailed, StateStarting, ErrInvalidStateTransition, StateFailed},
+		{"Shutdown to Stopped", StateShutdown, StateShutdown, StateStopped, ErrInvalidStateTransition, StateShutdown},
+		{"Shutdown to Starting", StateShutdown, StateShutdown, StateStarting, ErrInvalidStateTransition, StateShutdown},
+		{"Expected state mismatch", StateStopped, StateStarting, StateStarting, ErrExpectedStateMismatch, StateStopped},
 	}
 
 	for _, test := range tests {
 		t.Run(test.name, func(t *testing.T) {
-			p := &Process{
-				state: test.currentState,
-			}
+			p := NewProcess("test", 10, getTestSimpleResponderConfig("test"), debugLogger, debugLogger)
+			p.state = test.currentState
 
-			err := p.setState(test.newState)
+			resultState, err := p.swapState(test.expectedState, test.newState)
 			if err != nil && test.expectedError == nil {
 				t.Errorf("Unexpected error: %v", err)
 			} else if err == nil && test.expectedError != nil {
@@ -265,8 +280,8 @@ func TestSetState(t *testing.T) {
 				}
 			}
 
-			if p.state != test.expectedResult {
-				t.Errorf("Expected state: %v, got: %v", test.expectedResult, p.state)
+			if resultState != test.expectedResult {
+				t.Errorf("Expected state: %v, got: %v", test.expectedResult, resultState)
 			}
 		})
 	}
@@ -277,7 +292,6 @@ func TestProcess_ShutdownInterruptsHealthCheck(t *testing.T) {
 		t.Skip("skipping long shutdown test")
 	}
 
-	logMonitor := NewLogMonitorWriter(io.Discard)
 	expectedMessage := "testing91931"
 
 	// make a config where the healthcheck will always fail because port is wrong
@@ -285,13 +299,16 @@ func TestProcess_ShutdownInterruptsHealthCheck(t *testing.T) {
 	config.Proxy = "http://localhost:9998/test"
 
 	healthCheckTTLSeconds := 30
-	process := NewProcess("test-process", healthCheckTTLSeconds, config, logMonitor)
+	process := NewProcess("test-process", healthCheckTTLSeconds, config, debugLogger, debugLogger)
+
+	// make it a lot faster
+	process.healthCheckLoopInterval = time.Second
 
 	// start a goroutine to simulate a shutdown
 	var wg sync.WaitGroup
 	go func() {
 		defer wg.Done()
-		<-time.After(time.Second * 2)
+		<-time.After(time.Millisecond * 500)
 		process.Shutdown()
 	}()
 	wg.Add(1)
@@ -303,3 +320,23 @@ func TestProcess_ShutdownInterruptsHealthCheck(t *testing.T) {
 	assert.ErrorContains(t, err, "health check interrupted due to shutdown")
 	assert.Equal(t, StateShutdown, process.CurrentState())
 }
+
+func TestProcess_ExitInterruptsHealthCheck(t *testing.T) {
+	if testing.Short() {
+		t.Skip("skipping Exit Interrupts Health Check test")
+	}
+
+	// should run and exit but interrupt the long checkHealthTimeout
+	checkHealthTimeout := 5
+	config := ModelConfig{
+		Cmd:           "sleep 1",
+		Proxy:         "http://127.0.0.1:9913",
+		CheckEndpoint: "/health",
+	}
+
+	process := NewProcess("sleepy", checkHealthTimeout, config, debugLogger, debugLogger)
+	process.healthCheckLoopInterval = time.Second // make it faster
+	err := process.start()
+	assert.Equal(t, "upstream command exited prematurely but successfully", err.Error())
+	assert.Equal(t, process.CurrentState(), StateFailed)
+}

proxy/processgroup.go

@@ -0,0 +1,113 @@
+package proxy
+
+import (
+	"fmt"
+	"net/http"
+	"slices"
+	"sync"
+)
+
+type ProcessGroup struct {
+	sync.Mutex
+
+	config     Config
+	id         string
+	swap       bool
+	exclusive  bool
+	persistent bool
+
+	proxyLogger    *LogMonitor
+	upstreamLogger *LogMonitor
+
+	// map of current processes
+	processes       map[string]*Process
+	lastUsedProcess string
+}
+
+func NewProcessGroup(id string, config Config, proxyLogger *LogMonitor, upstreamLogger *LogMonitor) *ProcessGroup {
+	groupConfig, ok := config.Groups[id]
+	if !ok {
+		panic("Unable to find configuration for group id: " + id)
+	}
+
+	pg := &ProcessGroup{
+		id:             id,
+		config:         config,
+		swap:           groupConfig.Swap,
+		exclusive:      groupConfig.Exclusive,
+		persistent:     groupConfig.Persistent,
+		proxyLogger:    proxyLogger,
+		upstreamLogger: upstreamLogger,
+		processes:      make(map[string]*Process),
+	}
+
+	// Create a Process for each member in the group
+	for _, modelID := range groupConfig.Members {
+		modelConfig, modelID, _ := pg.config.FindConfig(modelID)
+		process := NewProcess(modelID, pg.config.HealthCheckTimeout, modelConfig, pg.upstreamLogger, pg.proxyLogger)
+		pg.processes[modelID] = process
+	}
+
+	return pg
+}
+
+// ProxyRequest proxies a request to the specified model
+func (pg *ProcessGroup) ProxyRequest(modelID string, writer http.ResponseWriter, request *http.Request) error {
+	if !pg.HasMember(modelID) {
+		return fmt.Errorf("model %s not part of group %s", modelID, pg.id)
+	}
+
+	if pg.swap {
+		pg.Lock()
+		if pg.lastUsedProcess != modelID {
+			if pg.lastUsedProcess != "" {
+				pg.processes[pg.lastUsedProcess].Stop()
+			}
+			pg.lastUsedProcess = modelID
+		}
+		pg.Unlock()
+	}
+
+	pg.processes[modelID].ProxyRequest(writer, request)
+	return nil
+}
+
+func (pg *ProcessGroup) HasMember(modelName string) bool {
+	return slices.Contains(pg.config.Groups[pg.id].Members, modelName)
+}
+
+func (pg *ProcessGroup) StopProcesses() {
+	pg.Lock()
+	defer pg.Unlock()
+	pg.stopProcesses()
+}
+
+// stopProcesses stops all processes in the group
+func (pg *ProcessGroup) stopProcesses() {
+	if len(pg.processes) == 0 {
+		return
+	}
+
+	// stop Processes in parallel
+	var wg sync.WaitGroup
+	for _, process := range pg.processes {
+		wg.Add(1)
+		go func(process *Process) {
+			defer wg.Done()
+			process.Stop()
+		}(process)
+	}
+	wg.Wait()
+}
+
+func (pg *ProcessGroup) Shutdown() {
+	var wg sync.WaitGroup
+	for _, process := range pg.processes {
+		wg.Add(1)
+		go func(process *Process) {
+			defer wg.Done()
+			process.Shutdown()
+		}(process)
+	}
+	wg.Wait()
+}

proxy/processgroup_test.go

@@ -0,0 +1,96 @@
+package proxy
+
+import (
+	"bytes"
+	"net/http"
+	"net/http/httptest"
+	"testing"
+
+	"github.com/stretchr/testify/assert"
+)
+
+var processGroupTestConfig = AddDefaultGroupToConfig(Config{
+	HealthCheckTimeout: 15,
+	Models: map[string]ModelConfig{
+		"model1": getTestSimpleResponderConfig("model1"),
+		"model2": getTestSimpleResponderConfig("model2"),
+		"model3": getTestSimpleResponderConfig("model3"),
+		"model4": getTestSimpleResponderConfig("model4"),
+		"model5": getTestSimpleResponderConfig("model5"),
+	},
+	Groups: map[string]GroupConfig{
+		"G1": {
+			Swap:      true,
+			Exclusive: true,
+			Members:   []string{"model1", "model2"},
+		},
+		"G2": {
+			Swap:      false,
+			Exclusive: true,
+			Members:   []string{"model3", "model4"},
+		},
+	},
+})
+
+func TestProcessGroup_DefaultHasCorrectModel(t *testing.T) {
+	pg := NewProcessGroup(DEFAULT_GROUP_ID, processGroupTestConfig, testLogger, testLogger)
+	assert.True(t, pg.HasMember("model5"))
+}
+
+func TestProcessGroup_HasMember(t *testing.T) {
+	pg := NewProcessGroup("G1", processGroupTestConfig, testLogger, testLogger)
+	assert.True(t, pg.HasMember("model1"))
+	assert.True(t, pg.HasMember("model2"))
+	assert.False(t, pg.HasMember("model3"))
+}
+
+func TestProcessGroup_ProxyRequestSwapIsTrue(t *testing.T) {
+	pg := NewProcessGroup("G1", processGroupTestConfig, testLogger, testLogger)
+	defer pg.StopProcesses()
+
+	tests := []string{"model1", "model2"}
+
+	for _, modelName := range tests {
+		t.Run(modelName, func(t *testing.T) {
+			reqBody := `{"x", "y"}`
+			req := httptest.NewRequest("POST", "/v1/chat/completions", bytes.NewBufferString(reqBody))
+			w := httptest.NewRecorder()
+
+			assert.NoError(t, pg.ProxyRequest(modelName, w, req))
+			assert.Equal(t, http.StatusOK, w.Code)
+			assert.Contains(t, w.Body.String(), modelName)
+
+			// make sure only one process is in the running state
+			count := 0
+			for _, process := range pg.processes {
+				if process.CurrentState() == StateReady {
+					count++
+				}
+			}
+			assert.Equal(t, 1, count)
+		})
+	}
+}
+
+func TestProcessGroup_ProxyRequestSwapIsFalse(t *testing.T) {
+	pg := NewProcessGroup("G2", processGroupTestConfig, testLogger, testLogger)
+	defer pg.StopProcesses()
+
+	tests := []string{"model3", "model4"}
+
+	for _, modelName := range tests {
+		t.Run(modelName, func(t *testing.T) {
+			reqBody := `{"x", "y"}`
+			req := httptest.NewRequest("POST", "/v1/chat/completions", bytes.NewBufferString(reqBody))
+			w := httptest.NewRecorder()
+			assert.NoError(t, pg.ProxyRequest(modelName, w, req))
+			assert.Equal(t, http.StatusOK, w.Code)
+			assert.Contains(t, w.Body.String(), modelName)
+		})
+	}
+
+	// make sure all the processes are running
+	for _, process := range pg.processes {
+		assert.Equal(t, StateReady, process.CurrentState())
+	}
+}

proxy/proxymanager.go

@@ -5,7 +5,9 @@ import (
 	"encoding/json"
 	"fmt"
 	"io"
+	"mime/multipart"
 	"net/http"
+	"os"
 	"sort"
 	"strconv"
 	"strings"
@@ -24,61 +26,111 @@ const (
 type ProxyManager struct {
 	sync.Mutex
 
-	config           *Config
-	currentProcesses map[string]*Process
-	logMonitor       *LogMonitor
-	ginEngine        *gin.Engine
+	config    Config
+	ginEngine *gin.Engine
+
+	// logging
+	proxyLogger    *LogMonitor
+	upstreamLogger *LogMonitor
+	muxLogger      *LogMonitor
+
+	processGroups map[string]*ProcessGroup
 }
 
-func New(config *Config) *ProxyManager {
-	pm := &ProxyManager{
-		config:           config,
-		currentProcesses: make(map[string]*Process),
-		logMonitor:       NewLogMonitor(),
-		ginEngine:        gin.New(),
-	}
+func New(config Config) *ProxyManager {
+	// set up loggers
+	stdoutLogger := NewLogMonitorWriter(os.Stdout)
+	upstreamLogger := NewLogMonitorWriter(stdoutLogger)
+	proxyLogger := NewLogMonitorWriter(stdoutLogger)
 
 	if config.LogRequests {
-		pm.ginEngine.Use(func(c *gin.Context) {
-			// Start timer
-			start := time.Now()
-
-			// capture these because /upstream/:model rewrites them in c.Next()
-			clientIP := c.ClientIP()
-			method := c.Request.Method
-			path := c.Request.URL.Path
-
-			// Process request
-			c.Next()
-
-			// Stop timer
-			duration := time.Since(start)
-
-			statusCode := c.Writer.Status()
-			bodySize := c.Writer.Size()
-
-			fmt.Fprintf(pm.logMonitor, "[llama-swap] %s [%s] \"%s %s %s\" %d %d \"%s\" %v\n",
-				clientIP,
-				time.Now().Format("2006-01-02 15:04:05"),
-				method,
-				path,
-				c.Request.Proto,
-				statusCode,
-				bodySize,
-				c.Request.UserAgent(),
-				duration,
-			)
-		})
+		proxyLogger.Warn("LogRequests configuration is deprecated. Use logLevel instead.")
 	}
 
-	// see: https://github.com/mostlygeek/llama-swap/issues/42
+	switch strings.ToLower(strings.TrimSpace(config.LogLevel)) {
+	case "debug":
+		proxyLogger.SetLogLevel(LevelDebug)
+		upstreamLogger.SetLogLevel(LevelDebug)
+	case "info":
+		proxyLogger.SetLogLevel(LevelInfo)
+		upstreamLogger.SetLogLevel(LevelInfo)
+	case "warn":
+		proxyLogger.SetLogLevel(LevelWarn)
+		upstreamLogger.SetLogLevel(LevelWarn)
+	case "error":
+		proxyLogger.SetLogLevel(LevelError)
+		upstreamLogger.SetLogLevel(LevelError)
+	default:
+		proxyLogger.SetLogLevel(LevelInfo)
+		upstreamLogger.SetLogLevel(LevelInfo)
+	}
+
+	pm := &ProxyManager{
+		config:    config,
+		ginEngine: gin.New(),
+
+		proxyLogger:    proxyLogger,
+		muxLogger:      stdoutLogger,
+		upstreamLogger: upstreamLogger,
+
+		processGroups: make(map[string]*ProcessGroup),
+	}
+
+	// create the process groups
+	for groupID := range config.Groups {
+		processGroup := NewProcessGroup(groupID, config, proxyLogger, upstreamLogger)
+		pm.processGroups[groupID] = processGroup
+	}
+
+	pm.ginEngine.Use(func(c *gin.Context) {
+		// Start timer
+		start := time.Now()
+
+		// capture these because /upstream/:model rewrites them in c.Next()
+		clientIP := c.ClientIP()
+		method := c.Request.Method
+		path := c.Request.URL.Path
+
+		// Process request
+		c.Next()
+
+		// Stop timer
+		duration := time.Since(start)
+
+		statusCode := c.Writer.Status()
+		bodySize := c.Writer.Size()
+
+		pm.proxyLogger.Infof("Request %s \"%s %s %s\" %d %d \"%s\" %v",
+			clientIP,
+			method,
+			path,
+			c.Request.Proto,
+			statusCode,
+			bodySize,
+			c.Request.UserAgent(),
+			duration,
+		)
+	})
+
+	// see: issue: #81, #77 and #42 for CORS issues
 	// respond with permissive OPTIONS for any endpoint
 	pm.ginEngine.Use(func(c *gin.Context) {
 		if c.Request.Method == "OPTIONS" {
 			c.Header("Access-Control-Allow-Origin", "*")
-			c.Header("Access-Control-Allow-Methods", "GET, POST, OPTIONS")
-			c.Header("Access-Control-Allow-Headers", "Content-Type, Authorization")
-			c.AbortWithStatus(204)
+			c.Header("Access-Control-Allow-Methods", "GET, POST, PUT, PATCH, DELETE, OPTIONS")
+
+			// allow whatever the client requested by default
+			if headers := c.Request.Header.Get("Access-Control-Request-Headers"); headers != "" {
+				sanitized := SanitizeAccessControlRequestHeaderValues(headers)
+				c.Header("Access-Control-Allow-Headers", sanitized)
+			} else {
+				c.Header(
+					"Access-Control-Allow-Headers",
+					"Content-Type, Authorization, Accept, X-Requested-With",
+				)
+			}
+			c.Header("Access-Control-Max-Age", "86400")
+			c.AbortWithStatus(http.StatusNoContent)
 			return
 		}
 		c.Next()
@@ -95,6 +147,7 @@ func New(config *Config) *ProxyManager {
 
 	// Support audio/speech endpoint
 	pm.ginEngine.POST("/v1/audio/speech", pm.proxyOAIHandler)
+	pm.ginEngine.POST("/v1/audio/transcriptions", pm.proxyOAIPostFormHandler)
 
 	pm.ginEngine.GET("/v1/models", pm.listModelsHandler)
 
@@ -102,12 +155,16 @@ func New(config *Config) *ProxyManager {
 	pm.ginEngine.GET("/logs", pm.sendLogsHandlers)
 	pm.ginEngine.GET("/logs/stream", pm.streamLogsHandler)
 	pm.ginEngine.GET("/logs/streamSSE", pm.streamLogsHandlerSSE)
+	pm.ginEngine.GET("/logs/stream/:logMonitorID", pm.streamLogsHandler)
+	pm.ginEngine.GET("/logs/streamSSE/:logMonitorID", pm.streamLogsHandlerSSE)
 
 	pm.ginEngine.GET("/upstream", pm.upstreamIndex)
 	pm.ginEngine.Any("/upstream/:model_id/*upstreamPath", pm.proxyToUpstream)
 
 	pm.ginEngine.GET("/unload", pm.unloadAllModelsHandler)
 
+	pm.ginEngine.GET("/running", pm.listRunningProcessesHandler)
+
 	pm.ginEngine.GET("/", func(c *gin.Context) {
 		// Set the Content-Type header to text/html
 		c.Header("Content-Type", "text/html")
@@ -151,27 +208,17 @@ func (pm *ProxyManager) StopProcesses() {
 	pm.Lock()
 	defer pm.Unlock()
 
-	pm.stopProcesses()
-}
-
-// for internal usage
-func (pm *ProxyManager) stopProcesses() {
-	if len(pm.currentProcesses) == 0 {
-		return
-	}
-
 	// stop Processes in parallel
 	var wg sync.WaitGroup
-	for _, process := range pm.currentProcesses {
+	for _, processGroup := range pm.processGroups {
 		wg.Add(1)
-		go func(process *Process) {
+		go func(processGroup *ProcessGroup) {
 			defer wg.Done()
-			process.Stop()
-		}(process)
+			processGroup.stopProcesses()
+		}(processGroup)
 	}
-	wg.Wait()
 
-	pm.currentProcesses = make(map[string]*Process)
+	wg.Wait()
 }
 
 // Shutdown is called to shutdown all upstream processes
@@ -180,18 +227,44 @@ func (pm *ProxyManager) Shutdown() {
 	pm.Lock()
 	defer pm.Unlock()
 
-	// shutdown process in parallel
+	pm.proxyLogger.Debug("Shutdown() called in proxy manager")
+
 	var wg sync.WaitGroup
-	for _, process := range pm.currentProcesses {
+	// Send shutdown signal to all process in groups
+	for _, processGroup := range pm.processGroups {
 		wg.Add(1)
-		go func(process *Process) {
+		go func(processGroup *ProcessGroup) {
 			defer wg.Done()
-			process.Shutdown()
-		}(process)
+			processGroup.Shutdown()
+		}(processGroup)
 	}
 	wg.Wait()
 }
 
+func (pm *ProxyManager) swapProcessGroup(requestedModel string) (*ProcessGroup, string, error) {
+	// de-alias the real model name and get a real one
+	realModelName, found := pm.config.RealModelName(requestedModel)
+	if !found {
+		return nil, realModelName, fmt.Errorf("could not find real modelID for %s", requestedModel)
+	}
+
+	processGroup := pm.findGroupByModelName(realModelName)
+	if processGroup == nil {
+		return nil, realModelName, fmt.Errorf("could not find process group for model %s", requestedModel)
+	}
+
+	if processGroup.exclusive {
+		pm.proxyLogger.Debugf("Exclusive mode for group %s, stopping other process groups", processGroup.id)
+		for groupId, otherGroup := range pm.processGroups {
+			if groupId != processGroup.id && !otherGroup.persistent {
+				otherGroup.StopProcesses()
+			}
+		}
+	}
+
+	return processGroup, realModelName, nil
+}
+
 func (pm *ProxyManager) listModelsHandler(c *gin.Context) {
 	data := []interface{}{}
 	for id, modelConfig := range pm.config.Models {
@@ -221,78 +294,6 @@ func (pm *ProxyManager) listModelsHandler(c *gin.Context) {
 	}
 }
 
-func (pm *ProxyManager) swapModel(requestedModel string) (*Process, error) {
-	pm.Lock()
-	defer pm.Unlock()
-
-	// Check if requestedModel contains a PROFILE_SPLIT_CHAR
-	profileName, modelName := splitRequestedModel(requestedModel)
-
-	if profileName != "" {
-		if _, found := pm.config.Profiles[profileName]; !found {
-			return nil, fmt.Errorf("model group not found %s", profileName)
-		}
-	}
-
-	// de-alias the real model name and get a real one
-	realModelName, found := pm.config.RealModelName(modelName)
-	if !found {
-		return nil, fmt.Errorf("could not find modelID for %s", requestedModel)
-	}
-
-	// check if model is part of the profile
-	if profileName != "" {
-		found := false
-		for _, item := range pm.config.Profiles[profileName] {
-			if item == realModelName {
-				found = true
-				break
-			}
-		}
-
-		if !found {
-			return nil, fmt.Errorf("model %s part of profile %s", realModelName, profileName)
-		}
-	}
-
-	// exit early when already running, otherwise stop everything and swap
-	requestedProcessKey := ProcessKeyName(profileName, realModelName)
-
-	if process, found := pm.currentProcesses[requestedProcessKey]; found {
-		return process, nil
-	}
-
-	// stop all running models
-	pm.stopProcesses()
-
-	if profileName == "" {
-		modelConfig, modelID, found := pm.config.FindConfig(realModelName)
-		if !found {
-			return nil, fmt.Errorf("could not find configuration for %s", realModelName)
-		}
-
-		process := NewProcess(modelID, pm.config.HealthCheckTimeout, modelConfig, pm.logMonitor)
-		processKey := ProcessKeyName(profileName, modelID)
-		pm.currentProcesses[processKey] = process
-	} else {
-		for _, modelName := range pm.config.Profiles[profileName] {
-			if realModelName, found := pm.config.RealModelName(modelName); found {
-				modelConfig, modelID, found := pm.config.FindConfig(realModelName)
-				if !found {
-					return nil, fmt.Errorf("could not find configuration for %s in group %s", realModelName, profileName)
-				}
-
-				process := NewProcess(modelID, pm.config.HealthCheckTimeout, modelConfig, pm.logMonitor)
-				processKey := ProcessKeyName(profileName, modelID)
-				pm.currentProcesses[processKey] = process
-			}
-		}
-	}
-
-	// requestedProcessKey should exist due to swap
-	return pm.currentProcesses[requestedProcessKey], nil
-}
-
 func (pm *ProxyManager) proxyToUpstream(c *gin.Context) {
 	requestedModel := c.Param("model_id")
 
@@ -301,13 +302,14 @@ func (pm *ProxyManager) proxyToUpstream(c *gin.Context) {
 		return
 	}
 
-	if process, err := pm.swapModel(requestedModel); err != nil {
-		pm.sendErrorResponse(c, http.StatusNotFound, fmt.Sprintf("unable to swap to model, %s", err.Error()))
-	} else {
-		// rewrite the path
-		c.Request.URL.Path = c.Param("upstreamPath")
-		process.ProxyRequest(c.Writer, c.Request)
+	processGroup, _, err := pm.swapProcessGroup(requestedModel)
+	if err != nil {
+		pm.sendErrorResponse(c, http.StatusInternalServerError, fmt.Sprintf("error swapping process group: %s", err.Error()))
 	}
+
+	// rewrite the path
+	c.Request.URL.Path = c.Param("upstreamPath")
+	processGroup.ProxyRequest(requestedModel, c.Writer, c.Request)
 }
 
 func (pm *ProxyManager) upstreamIndex(c *gin.Context) {
@@ -347,28 +349,134 @@ func (pm *ProxyManager) proxyOAIHandler(c *gin.Context) {
 		pm.sendErrorResponse(c, http.StatusBadRequest, "missing or invalid 'model' key")
 	}
 
-	if process, err := pm.swapModel(requestedModel); err != nil {
-		pm.sendErrorResponse(c, http.StatusNotFound, fmt.Sprintf("unable to swap to model, %s", err.Error()))
-		return
-	} else {
+	processGroup, realModelName, err := pm.swapProcessGroup(requestedModel)
+	if err != nil {
+		pm.sendErrorResponse(c, http.StatusInternalServerError, fmt.Sprintf("error swapping process group: %s", err.Error()))
+	}
 
-		// strip
-		profileName, modelName := splitRequestedModel(requestedModel)
-		if profileName != "" {
-			bodyBytes, err = sjson.SetBytes(bodyBytes, "model", modelName)
+	// issue #69 allow custom model names to be sent to upstream
+	useModelName := pm.config.Models[realModelName].UseModelName
+	if useModelName != "" {
+		bodyBytes, err = sjson.SetBytes(bodyBytes, "model", useModelName)
+		if err != nil {
+			pm.sendErrorResponse(c, http.StatusInternalServerError, fmt.Sprintf("error rewriting model name in JSON: %s", err.Error()))
+			return
+		}
+	}
+
+	c.Request.Body = io.NopCloser(bytes.NewBuffer(bodyBytes))
+
+	// dechunk it as we already have all the body bytes see issue #11
+	c.Request.Header.Del("transfer-encoding")
+	c.Request.Header.Add("content-length", strconv.Itoa(len(bodyBytes)))
+
+	if err := processGroup.ProxyRequest(realModelName, c.Writer, c.Request); err != nil {
+		pm.sendErrorResponse(c, http.StatusInternalServerError, fmt.Sprintf("error proxying request: %s", err.Error()))
+		pm.proxyLogger.Errorf("Error Proxying Request for processGroup %s and model %s", processGroup.id, realModelName)
+	}
+}
+
+func (pm *ProxyManager) proxyOAIPostFormHandler(c *gin.Context) {
+	// We need to reconstruct the multipart form in any case since the body is consumed
+	// Create a new buffer for the reconstructed request
+	var requestBuffer bytes.Buffer
+	multipartWriter := multipart.NewWriter(&requestBuffer)
+
+	// Parse multipart form
+	if err := c.Request.ParseMultipartForm(32 << 20); err != nil { // 32MB max memory, larger files go to tmp disk
+		pm.sendErrorResponse(c, http.StatusBadRequest, fmt.Sprintf("error parsing multipart form: %s", err.Error()))
+		return
+	}
+
+	// Get model parameter from the form
+	requestedModel := c.Request.FormValue("model")
+	if requestedModel == "" {
+		pm.sendErrorResponse(c, http.StatusBadRequest, "missing or invalid 'model' parameter in form data")
+		return
+	}
+
+	processGroup, realModelName, err := pm.swapProcessGroup(requestedModel)
+	if err != nil {
+		pm.sendErrorResponse(c, http.StatusInternalServerError, fmt.Sprintf("error swapping process group: %s", err.Error()))
+	}
+
+	// Copy all form values
+	for key, values := range c.Request.MultipartForm.Value {
+		for _, value := range values {
+			fieldValue := value
+			// If this is the model field and we have a profile, use just the model name
+			if key == "model" {
+				// # issue #69 allow custom model names to be sent to upstream
+				useModelName := pm.config.Models[realModelName].UseModelName
+
+				if useModelName != "" {
+					fieldValue = useModelName
+				} else {
+					fieldValue = requestedModel
+				}
+			}
+			field, err := multipartWriter.CreateFormField(key)
 			if err != nil {
-				pm.sendErrorResponse(c, http.StatusInternalServerError, fmt.Sprintf("error updating JSON: %s", err.Error()))
+				pm.sendErrorResponse(c, http.StatusInternalServerError, "error recreating form field")
+				return
+			}
+			if _, err = field.Write([]byte(fieldValue)); err != nil {
+				pm.sendErrorResponse(c, http.StatusInternalServerError, "error writing form field")
 				return
 			}
 		}
+	}
 
-		c.Request.Body = io.NopCloser(bytes.NewBuffer(bodyBytes))
+	// Copy all files from the original request
+	for key, fileHeaders := range c.Request.MultipartForm.File {
+		for _, fileHeader := range fileHeaders {
+			formFile, err := multipartWriter.CreateFormFile(key, fileHeader.Filename)
+			if err != nil {
+				pm.sendErrorResponse(c, http.StatusInternalServerError, "error recreating form file")
+				return
+			}
 
-		// dechunk it as we already have all the body bytes see issue #11
-		c.Request.Header.Del("transfer-encoding")
-		c.Request.Header.Add("content-length", strconv.Itoa(len(bodyBytes)))
+			file, err := fileHeader.Open()
+			if err != nil {
+				pm.sendErrorResponse(c, http.StatusInternalServerError, "error opening uploaded file")
+				return
+			}
 
-		process.ProxyRequest(c.Writer, c.Request)
+			if _, err = io.Copy(formFile, file); err != nil {
+				file.Close()
+				pm.sendErrorResponse(c, http.StatusInternalServerError, "error copying file data")
+				return
+			}
+			file.Close()
+		}
+	}
+
+	// Close the multipart writer to finalize the form
+	if err := multipartWriter.Close(); err != nil {
+		pm.sendErrorResponse(c, http.StatusInternalServerError, "error finalizing multipart form")
+		return
+	}
+
+	// Create a new request with the reconstructed form data
+	modifiedReq, err := http.NewRequestWithContext(
+		c.Request.Context(),
+		c.Request.Method,
+		c.Request.URL.String(),
+		&requestBuffer,
+	)
+	if err != nil {
+		pm.sendErrorResponse(c, http.StatusInternalServerError, "error creating modified request")
+		return
+	}
+
+	// Copy the headers from the original request
+	modifiedReq.Header = c.Request.Header.Clone()
+	modifiedReq.Header.Set("Content-Type", multipartWriter.FormDataContentType())
+
+	// Use the modified request for proxying
+	if err := processGroup.ProxyRequest(realModelName, c.Writer, modifiedReq); err != nil {
+		pm.sendErrorResponse(c, http.StatusInternalServerError, fmt.Sprintf("error proxying request: %s", err.Error()))
+		pm.proxyLogger.Errorf("Error Proxying Request for processGroup %s and model %s", processGroup.id, realModelName)
 	}
 }
 
@@ -387,15 +495,34 @@ func (pm *ProxyManager) unloadAllModelsHandler(c *gin.Context) {
 	c.String(http.StatusOK, "OK")
 }
 
-func ProcessKeyName(groupName, modelName string) string {
-	return groupName + PROFILE_SPLIT_CHAR + modelName
+func (pm *ProxyManager) listRunningProcessesHandler(context *gin.Context) {
+	context.Header("Content-Type", "application/json")
+	runningProcesses := make([]gin.H, 0) // Default to an empty response.
+
+	for _, processGroup := range pm.processGroups {
+		for _, process := range processGroup.processes {
+			if process.CurrentState() == StateReady {
+				runningProcesses = append(runningProcesses, gin.H{
+					"model": process.ID,
+					"state": process.state,
+				})
+			}
+		}
+	}
+
+	// Put the results under the `running` key.
+	response := gin.H{
+		"running": runningProcesses,
+	}
+
+	context.JSON(http.StatusOK, response) // Always return 200 OK
 }
 
-func splitRequestedModel(requestedModel string) (string, string) {
-	profileName, modelName := "", requestedModel
-	if idx := strings.Index(requestedModel, PROFILE_SPLIT_CHAR); idx != -1 {
-		profileName = requestedModel[:idx]
-		modelName = requestedModel[idx+1:]
+func (pm *ProxyManager) findGroupByModelName(modelName string) *ProcessGroup {
+	for _, group := range pm.processGroups {
+		if group.HasMember(modelName) {
+			return group
+		}
 	}
-	return profileName, modelName
+	return nil
 }

proxy/proxymanager_loghandlers.go

@@ -9,7 +9,6 @@ import (
 )
 
 func (pm *ProxyManager) sendLogsHandlers(c *gin.Context) {
-
 	accept := c.GetHeader("Accept")
 	if strings.Contains(accept, "text/html") {
 		// Set the Content-Type header to text/html
@@ -28,7 +27,7 @@ func (pm *ProxyManager) sendLogsHandlers(c *gin.Context) {
 		}
 	} else {
 		c.Header("Content-Type", "text/plain")
-		history := pm.logMonitor.GetHistory()
+		history := pm.muxLogger.GetHistory()
 		_, err := c.Writer.Write(history)
 		if err != nil {
 			c.AbortWithError(http.StatusInternalServerError, err)
@@ -42,8 +41,14 @@ func (pm *ProxyManager) streamLogsHandler(c *gin.Context) {
 	c.Header("Transfer-Encoding", "chunked")
 	c.Header("X-Content-Type-Options", "nosniff")
 
-	ch := pm.logMonitor.Subscribe()
-	defer pm.logMonitor.Unsubscribe(ch)
+	logMonitorId := c.Param("logMonitorID")
+	logger, err := pm.getLogger(logMonitorId)
+	if err != nil {
+		c.String(http.StatusBadRequest, err.Error())
+		return
+	}
+	ch := logger.Subscribe()
+	defer logger.Unsubscribe(ch)
 
 	notify := c.Request.Context().Done()
 	flusher, ok := c.Writer.(http.Flusher)
@@ -56,7 +61,7 @@ func (pm *ProxyManager) streamLogsHandler(c *gin.Context) {
 	// Send history first if not skipped
 
 	if !skipHistory {
-		history := pm.logMonitor.GetHistory()
+		history := logger.GetHistory()
 		if len(history) != 0 {
 			c.Writer.Write(history)
 			flusher.Flush()
@@ -85,15 +90,21 @@ func (pm *ProxyManager) streamLogsHandlerSSE(c *gin.Context) {
 	c.Header("Connection", "keep-alive")
 	c.Header("X-Content-Type-Options", "nosniff")
 
-	ch := pm.logMonitor.Subscribe()
-	defer pm.logMonitor.Unsubscribe(ch)
+	logMonitorId := c.Param("logMonitorID")
+	logger, err := pm.getLogger(logMonitorId)
+	if err != nil {
+		c.String(http.StatusBadRequest, err.Error())
+		return
+	}
+	ch := logger.Subscribe()
+	defer logger.Unsubscribe(ch)
 
 	notify := c.Request.Context().Done()
 
 	// Send history first if not skipped
 	_, skipHistory := c.GetQuery("no-history")
 	if !skipHistory {
-		history := pm.logMonitor.GetHistory()
+		history := logger.GetHistory()
 		if len(history) != 0 {
 			c.SSEvent("message", string(history))
 			c.Writer.Flush()
@@ -111,3 +122,21 @@ func (pm *ProxyManager) streamLogsHandlerSSE(c *gin.Context) {
 		}
 	}
 }
+
+// getLogger searches for the appropriate logger based on the logMonitorId
+func (pm *ProxyManager) getLogger(logMonitorId string) (*LogMonitor, error) {
+	var logger *LogMonitor
+
+	if logMonitorId == "" {
+		// maintain the default
+		logger = pm.muxLogger
+	} else if logMonitorId == "proxy" {
+		logger = pm.proxyLogger
+	} else if logMonitorId == "upstream" {
+		logger = pm.upstreamLogger
+	} else {
+		return nil, fmt.Errorf("invalid logger. Use 'proxy' or 'upstream'")
+	}
+
+	return logger, nil
+}

proxy/proxymanager_test.go

@@ -4,6 +4,8 @@ import (
 	"bytes"
 	"encoding/json"
 	"fmt"
+	"math/rand"
+	"mime/multipart"
 	"net/http"
 	"net/http/httptest"
 	"sync"
@@ -14,13 +16,14 @@ import (
 )
 
 func TestProxyManager_SwapProcessCorrectly(t *testing.T) {
-	config := &Config{
+	config := AddDefaultGroupToConfig(Config{
 		HealthCheckTimeout: 15,
 		Models: map[string]ModelConfig{
 			"model1": getTestSimpleResponderConfig("model1"),
 			"model2": getTestSimpleResponderConfig("model2"),
 		},
-	}
+		LogLevel: "error",
+	})
 
 	proxy := New(config)
 	defer proxy.StopProcesses()
@@ -33,58 +36,91 @@ func TestProxyManager_SwapProcessCorrectly(t *testing.T) {
 		proxy.HandlerFunc(w, req)
 		assert.Equal(t, http.StatusOK, w.Code)
 		assert.Contains(t, w.Body.String(), modelName)
-
-		_, exists := proxy.currentProcesses[ProcessKeyName("", modelName)]
-		assert.True(t, exists, "expected %s key in currentProcesses", modelName)
-
 	}
-
-	// make sure there's only one loaded model
-	assert.Len(t, proxy.currentProcesses, 1)
 }
 
 func TestProxyManager_SwapMultiProcess(t *testing.T) {
-
-	model1 := "path1/model1"
-	model2 := "path2/model2"
-
-	profileModel1 := ProcessKeyName("test", model1)
-	profileModel2 := ProcessKeyName("test", model2)
-
-	config := &Config{
+	config := AddDefaultGroupToConfig(Config{
 		HealthCheckTimeout: 15,
 		Models: map[string]ModelConfig{
-			model1: getTestSimpleResponderConfig("model1"),
-			model2: getTestSimpleResponderConfig("model2"),
+			"model1": getTestSimpleResponderConfig("model1"),
+			"model2": getTestSimpleResponderConfig("model2"),
 		},
-		Profiles: map[string][]string{
-			"test": {model1, model2},
+		LogLevel: "error",
+		Groups: map[string]GroupConfig{
+			"G1": {
+				Swap:      true,
+				Exclusive: false,
+				Members:   []string{"model1"},
+			},
+			"G2": {
+				Swap:      true,
+				Exclusive: false,
+				Members:   []string{"model2"},
+			},
 		},
-	}
+	})
 
 	proxy := New(config)
 	defer proxy.StopProcesses()
 
-	for modelID, requestedModel := range map[string]string{
-		"model1": profileModel1,
-		"model2": profileModel2,
-	} {
+	tests := []string{"model1", "model2"}
+	for _, requestedModel := range tests {
+		t.Run(requestedModel, func(t *testing.T) {
+			reqBody := fmt.Sprintf(`{"model":"%s"}`, requestedModel)
+			req := httptest.NewRequest("POST", "/v1/chat/completions", bytes.NewBufferString(reqBody))
+			w := httptest.NewRecorder()
+
+			proxy.HandlerFunc(w, req)
+			assert.Equal(t, http.StatusOK, w.Code)
+			assert.Contains(t, w.Body.String(), requestedModel)
+
+		})
+	}
+
+	// make sure there's two loaded models
+	assert.Equal(t, proxy.findGroupByModelName("model1").processes["model1"].CurrentState(), StateReady)
+	assert.Equal(t, proxy.findGroupByModelName("model2").processes["model2"].CurrentState(), StateReady)
+}
+
+// Test that a persistent group is not affected by the swapping behaviour of
+// other groups.
+func TestProxyManager_PersistentGroupsAreNotSwapped(t *testing.T) {
+	config := AddDefaultGroupToConfig(Config{
+		HealthCheckTimeout: 15,
+		Models: map[string]ModelConfig{
+			"model1": getTestSimpleResponderConfig("model1"), // goes into the default group
+			"model2": getTestSimpleResponderConfig("model2"),
+		},
+		LogLevel: "error",
+		Groups: map[string]GroupConfig{
+			// the forever group is persistent and should not be affected by model1
+			"forever": {
+				Swap:       true,
+				Exclusive:  false,
+				Persistent: true,
+				Members:    []string{"model2"},
+			},
+		},
+	})
+
+	proxy := New(config)
+	defer proxy.StopProcesses()
+
+	// make requests to load all models, loading model1 should not affect model2
+	tests := []string{"model2", "model1"}
+	for _, requestedModel := range tests {
 		reqBody := fmt.Sprintf(`{"model":"%s"}`, requestedModel)
 		req := httptest.NewRequest("POST", "/v1/chat/completions", bytes.NewBufferString(reqBody))
 		w := httptest.NewRecorder()
 
 		proxy.HandlerFunc(w, req)
 		assert.Equal(t, http.StatusOK, w.Code)
-		assert.Contains(t, w.Body.String(), modelID)
+		assert.Contains(t, w.Body.String(), requestedModel)
 	}
 
-	// make sure there's two loaded models
-	assert.Len(t, proxy.currentProcesses, 2)
-	_, exists := proxy.currentProcesses[profileModel1]
-	assert.True(t, exists, "expected "+profileModel1+" key in currentProcesses")
-
-	_, exists = proxy.currentProcesses[profileModel2]
-	assert.True(t, exists, "expected "+profileModel2+" key in currentProcesses")
+	assert.Equal(t, proxy.findGroupByModelName("model2").processes["model2"].CurrentState(), StateReady)
+	assert.Equal(t, proxy.findGroupByModelName("model1").processes["model1"].CurrentState(), StateReady)
 }
 
 // When a request for a different model comes in ProxyManager should wait until
@@ -94,14 +130,15 @@ func TestProxyManager_SwapMultiProcessParallelRequests(t *testing.T) {
 		t.Skip("skipping slow test")
 	}
 
-	config := &Config{
+	config := AddDefaultGroupToConfig(Config{
 		HealthCheckTimeout: 15,
 		Models: map[string]ModelConfig{
 			"model1": getTestSimpleResponderConfig("model1"),
 			"model2": getTestSimpleResponderConfig("model2"),
 			"model3": getTestSimpleResponderConfig("model3"),
 		},
-	}
+		LogLevel: "error",
+	})
 
 	proxy := New(config)
 	defer proxy.StopProcesses()
@@ -144,13 +181,14 @@ func TestProxyManager_SwapMultiProcessParallelRequests(t *testing.T) {
 }
 
 func TestProxyManager_ListModelsHandler(t *testing.T) {
-	config := &Config{
+	config := Config{
 		HealthCheckTimeout: 15,
 		Models: map[string]ModelConfig{
 			"model1": getTestSimpleResponderConfig("model1"),
 			"model2": getTestSimpleResponderConfig("model2"),
 			"model3": getTestSimpleResponderConfig("model3"),
 		},
+		LogLevel: "error",
 	}
 
 	proxy := New(config)
@@ -211,50 +249,6 @@ func TestProxyManager_ListModelsHandler(t *testing.T) {
 	assert.Empty(t, expectedModels, "not all expected models were returned")
 }
 
-func TestProxyManager_ProfileNonMember(t *testing.T) {
-
-	model1 := "path1/model1"
-	model2 := "path2/model2"
-
-	profileMemberName := ProcessKeyName("test", model1)
-	profileNonMemberName := ProcessKeyName("test", model2)
-
-	config := &Config{
-		HealthCheckTimeout: 15,
-		Models: map[string]ModelConfig{
-			model1: getTestSimpleResponderConfig("model1"),
-			model2: getTestSimpleResponderConfig("model2"),
-		},
-		Profiles: map[string][]string{
-			"test": {model1},
-		},
-	}
-
-	proxy := New(config)
-	defer proxy.StopProcesses()
-
-	// actual member of profile
-	{
-		reqBody := fmt.Sprintf(`{"model":"%s"}`, profileMemberName)
-		req := httptest.NewRequest("POST", "/v1/chat/completions", bytes.NewBufferString(reqBody))
-		w := httptest.NewRecorder()
-
-		proxy.HandlerFunc(w, req)
-		assert.Equal(t, http.StatusOK, w.Code)
-		assert.Contains(t, w.Body.String(), "model1")
-	}
-
-	// actual model, but non-member will 404
-	{
-		reqBody := fmt.Sprintf(`{"model":"%s"}`, profileNonMemberName)
-		req := httptest.NewRequest("POST", "/v1/chat/completions", bytes.NewBufferString(reqBody))
-		w := httptest.NewRecorder()
-
-		proxy.HandlerFunc(w, req)
-		assert.Equal(t, http.StatusNotFound, w.Code)
-	}
-}
-
 func TestProxyManager_Shutdown(t *testing.T) {
 	// make broken model configurations
 	model1Config := getTestSimpleResponderConfigPort("model1", 9991)
@@ -266,23 +260,27 @@ func TestProxyManager_Shutdown(t *testing.T) {
 	model3Config := getTestSimpleResponderConfigPort("model3", 9993)
 	model3Config.Proxy = "http://localhost:10003/"
 
-	config := &Config{
+	config := AddDefaultGroupToConfig(Config{
 		HealthCheckTimeout: 15,
-		Profiles: map[string][]string{
-			"test": {"model1", "model2", "model3"},
-		},
 		Models: map[string]ModelConfig{
 			"model1": model1Config,
 			"model2": model2Config,
 			"model3": model3Config,
 		},
-	}
+		LogLevel: "error",
+		Groups: map[string]GroupConfig{
+			"test": {
+				Swap:    false,
+				Members: []string{"model1", "model2", "model3"},
+			},
+		},
+	})
 
 	proxy := New(config)
 
 	// Start all the processes
 	var wg sync.WaitGroup
-	for _, modelName := range []string{"test:model1", "test:model2", "test:model3"} {
+	for _, modelName := range []string{"model1", "model2", "model3"} {
 		wg.Add(1)
 		go func(modelName string) {
 			defer wg.Done()
@@ -290,11 +288,10 @@ func TestProxyManager_Shutdown(t *testing.T) {
 			req := httptest.NewRequest("POST", "/v1/chat/completions", bytes.NewBufferString(reqBody))
 			w := httptest.NewRecorder()
 
-			// send a request to trigger the proxy to load
+			// send a request to trigger the proxy to load ... this should hang waiting for start up
 			proxy.HandlerFunc(w, req)
 			assert.Equal(t, http.StatusBadGateway, w.Code)
 			assert.Contains(t, w.Body.String(), "health check interrupted due to shutdown")
-			//fmt.Println(w.Code, w.Body.String())
 		}(modelName)
 	}
 
@@ -306,46 +303,292 @@ func TestProxyManager_Shutdown(t *testing.T) {
 }
 
 func TestProxyManager_Unload(t *testing.T) {
-	config := &Config{
+	config := AddDefaultGroupToConfig(Config{
 		HealthCheckTimeout: 15,
 		Models: map[string]ModelConfig{
 			"model1": getTestSimpleResponderConfig("model1"),
 		},
-	}
+		LogLevel: "error",
+	})
 
 	proxy := New(config)
-	proc, err := proxy.swapModel("model1")
-	assert.NoError(t, err)
-	assert.NotNil(t, proc)
-
-	assert.Len(t, proxy.currentProcesses, 1)
-	req := httptest.NewRequest("GET", "/unload", nil)
+	reqBody := fmt.Sprintf(`{"model":"%s"}`, "model1")
+	req := httptest.NewRequest("POST", "/v1/chat/completions", bytes.NewBufferString(reqBody))
 	w := httptest.NewRecorder()
 	proxy.HandlerFunc(w, req)
+
+	assert.Equal(t, proxy.processGroups[DEFAULT_GROUP_ID].processes["model1"].CurrentState(), StateReady)
+	req = httptest.NewRequest("GET", "/unload", nil)
+	w = httptest.NewRecorder()
+	proxy.HandlerFunc(w, req)
 	assert.Equal(t, http.StatusOK, w.Code)
 	assert.Equal(t, w.Body.String(), "OK")
-	assert.Len(t, proxy.currentProcesses, 0)
+
+	// give it a bit of time to stop
+	<-time.After(time.Millisecond * 250)
+	assert.Equal(t, proxy.processGroups[DEFAULT_GROUP_ID].processes["model1"].CurrentState(), StateStopped)
 }
 
-// issue 62, strip profile slug from model name
-func TestProxyManager_StripProfileSlug(t *testing.T) {
-	config := &Config{
+// Test issue #61 `Listing the current list of models and the loaded model.`
+func TestProxyManager_RunningEndpoint(t *testing.T) {
+
+	// Shared configuration
+	config := AddDefaultGroupToConfig(Config{
 		HealthCheckTimeout: 15,
-		Profiles: map[string][]string{
-			"test": {"TheExpectedModel"}, // TheExpectedModel is default in simple-responder.go
+		Models: map[string]ModelConfig{
+			"model1": getTestSimpleResponderConfig("model1"),
+			"model2": getTestSimpleResponderConfig("model2"),
 		},
+		LogLevel: "debug",
+	})
+
+	// Define a helper struct to parse the JSON response.
+	type RunningResponse struct {
+		Running []struct {
+			Model string `json:"model"`
+			State string `json:"state"`
+		} `json:"running"`
+	}
+
+	// Create proxy once for all tests
+	proxy := New(config)
+	defer proxy.StopProcesses()
+
+	t.Run("no models loaded", func(t *testing.T) {
+		req := httptest.NewRequest("GET", "/running", nil)
+		w := httptest.NewRecorder()
+		proxy.HandlerFunc(w, req)
+
+		assert.Equal(t, http.StatusOK, w.Code)
+
+		var response RunningResponse
+
+		// Check if this is a valid JSON object.
+		assert.NoError(t, json.Unmarshal(w.Body.Bytes(), &response))
+
+		// We should have an empty running array here.
+		assert.Empty(t, response.Running, "expected no running models")
+	})
+
+	t.Run("single model loaded", func(t *testing.T) {
+		// Load just a model.
+		reqBody := `{"model":"model1"}`
+		req := httptest.NewRequest("POST", "/v1/chat/completions", bytes.NewBufferString(reqBody))
+		w := httptest.NewRecorder()
+		proxy.HandlerFunc(w, req)
+		assert.Equal(t, http.StatusOK, w.Code)
+
+		// Simulate browser call for the `/running` endpoint.
+		req = httptest.NewRequest("GET", "/running", nil)
+		w = httptest.NewRecorder()
+		proxy.HandlerFunc(w, req)
+
+		var response RunningResponse
+		assert.NoError(t, json.Unmarshal(w.Body.Bytes(), &response))
+
+		// Check if we have a single array element.
+		assert.Len(t, response.Running, 1)
+
+		// Is this the right model?
+		assert.Equal(t, "model1", response.Running[0].Model)
+
+		// Is the model loaded?
+		assert.Equal(t, "ready", response.Running[0].State)
+	})
+}
+
+func TestProxyManager_AudioTranscriptionHandler(t *testing.T) {
+	config := AddDefaultGroupToConfig(Config{
+		HealthCheckTimeout: 15,
 		Models: map[string]ModelConfig{
 			"TheExpectedModel": getTestSimpleResponderConfig("TheExpectedModel"),
 		},
-	}
+		LogLevel: "error",
+	})
 
 	proxy := New(config)
 	defer proxy.StopProcesses()
 
-	reqBody := fmt.Sprintf(`{"model":"%s"}`, "test:TheExpectedModel")
-	req := httptest.NewRequest("POST", "/v1/audio/speech", bytes.NewBufferString(reqBody))
-	w := httptest.NewRecorder()
-	proxy.HandlerFunc(w, req)
-	assert.Equal(t, http.StatusOK, w.Code)
-	assert.Contains(t, w.Body.String(), "ok")
+	// Create a buffer with multipart form data
+	var b bytes.Buffer
+	w := multipart.NewWriter(&b)
+
+	// Add the model field
+	fw, err := w.CreateFormField("model")
+	assert.NoError(t, err)
+	_, err = fw.Write([]byte("TheExpectedModel"))
+	assert.NoError(t, err)
+
+	// Add a file field
+	fw, err = w.CreateFormFile("file", "test.mp3")
+	assert.NoError(t, err)
+	// Generate random content length between 10 and 20
+	contentLength := rand.Intn(11) + 10 // 10 to 20
+	content := make([]byte, contentLength)
+	_, err = fw.Write(content)
+	assert.NoError(t, err)
+	w.Close()
+
+	// Create the request with the multipart form data
+	req := httptest.NewRequest("POST", "/v1/audio/transcriptions", &b)
+	req.Header.Set("Content-Type", w.FormDataContentType())
+	rec := httptest.NewRecorder()
+	proxy.HandlerFunc(rec, req)
+
+	// Verify the response
+	assert.Equal(t, http.StatusOK, rec.Code)
+	var response map[string]string
+	err = json.Unmarshal(rec.Body.Bytes(), &response)
+	assert.NoError(t, err)
+	assert.Equal(t, "TheExpectedModel", response["model"])
+	assert.Equal(t, response["text"], fmt.Sprintf("The length of the file is %d bytes", contentLength)) // matches simple-responder
+}
+
+// Test useModelName in configuration sends overrides what is sent to upstream
+func TestProxyManager_UseModelName(t *testing.T) {
+	upstreamModelName := "upstreamModel"
+
+	modelConfig := getTestSimpleResponderConfig(upstreamModelName)
+	modelConfig.UseModelName = upstreamModelName
+
+	config := AddDefaultGroupToConfig(Config{
+		HealthCheckTimeout: 15,
+		Models: map[string]ModelConfig{
+			"model1": modelConfig,
+		},
+		LogLevel: "error",
+	})
+
+	proxy := New(config)
+	defer proxy.StopProcesses()
+
+	requestedModel := "model1"
+
+	t.Run("useModelName over rides requested model: /v1/chat/completions", func(t *testing.T) {
+		reqBody := fmt.Sprintf(`{"model":"%s"}`, requestedModel)
+		req := httptest.NewRequest("POST", "/v1/chat/completions", bytes.NewBufferString(reqBody))
+		w := httptest.NewRecorder()
+
+		proxy.HandlerFunc(w, req)
+		assert.Equal(t, http.StatusOK, w.Code)
+		assert.Contains(t, w.Body.String(), upstreamModelName)
+	})
+
+	t.Run("useModelName over rides requested model: /v1/audio/transcriptions", func(t *testing.T) {
+		// Create a buffer with multipart form data
+		var b bytes.Buffer
+		w := multipart.NewWriter(&b)
+
+		// Add the model field
+		fw, err := w.CreateFormField("model")
+		assert.NoError(t, err)
+		_, err = fw.Write([]byte(requestedModel))
+		assert.NoError(t, err)
+
+		// Add a file field
+		fw, err = w.CreateFormFile("file", "test.mp3")
+		assert.NoError(t, err)
+		_, err = fw.Write([]byte("test"))
+		assert.NoError(t, err)
+		w.Close()
+
+		// Create the request with the multipart form data
+		req := httptest.NewRequest("POST", "/v1/audio/transcriptions", &b)
+		req.Header.Set("Content-Type", w.FormDataContentType())
+		rec := httptest.NewRecorder()
+		proxy.HandlerFunc(rec, req)
+
+		// Verify the response
+		assert.Equal(t, http.StatusOK, rec.Code)
+		var response map[string]string
+		err = json.Unmarshal(rec.Body.Bytes(), &response)
+		assert.NoError(t, err)
+		assert.Equal(t, upstreamModelName, response["model"])
+	})
+}
+
+func TestProxyManager_CORSOptionsHandler(t *testing.T) {
+	config := AddDefaultGroupToConfig(Config{
+		HealthCheckTimeout: 15,
+		Models: map[string]ModelConfig{
+			"model1": getTestSimpleResponderConfig("model1"),
+		},
+		LogLevel: "error",
+	})
+
+	tests := []struct {
+		name            string
+		method          string
+		requestHeaders  map[string]string
+		expectedStatus  int
+		expectedHeaders map[string]string
+	}{
+		{
+			name:           "OPTIONS with no headers",
+			method:         "OPTIONS",
+			expectedStatus: http.StatusNoContent,
+			expectedHeaders: map[string]string{
+				"Access-Control-Allow-Origin":  "*",
+				"Access-Control-Allow-Methods": "GET, POST, PUT, PATCH, DELETE, OPTIONS",
+				"Access-Control-Allow-Headers": "Content-Type, Authorization, Accept, X-Requested-With",
+			},
+		},
+		{
+			name:   "OPTIONS with specific headers",
+			method: "OPTIONS",
+			requestHeaders: map[string]string{
+				"Access-Control-Request-Headers": "X-Custom-Header, Some-Other-Header",
+			},
+			expectedStatus: http.StatusNoContent,
+			expectedHeaders: map[string]string{
+				"Access-Control-Allow-Origin":  "*",
+				"Access-Control-Allow-Methods": "GET, POST, PUT, PATCH, DELETE, OPTIONS",
+				"Access-Control-Allow-Headers": "X-Custom-Header, Some-Other-Header",
+			},
+		},
+		{
+			name:           "Non-OPTIONS request",
+			method:         "GET",
+			expectedStatus: http.StatusNotFound, // Since we don't have a GET route defined
+		},
+	}
+
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			proxy := New(config)
+			defer proxy.StopProcesses()
+
+			req := httptest.NewRequest(tt.method, "/v1/chat/completions", nil)
+			for k, v := range tt.requestHeaders {
+				req.Header.Set(k, v)
+			}
+
+			w := httptest.NewRecorder()
+			proxy.ginEngine.ServeHTTP(w, req)
+
+			assert.Equal(t, tt.expectedStatus, w.Code)
+
+			for header, expectedValue := range tt.expectedHeaders {
+				assert.Equal(t, expectedValue, w.Header().Get(header))
+			}
+		})
+	}
+}
+
+func TestProxyManager_Upstream(t *testing.T) {
+	config := AddDefaultGroupToConfig(Config{
+		HealthCheckTimeout: 15,
+		Models: map[string]ModelConfig{
+			"model1": getTestSimpleResponderConfig("model1"),
+		},
+		LogLevel: "error",
+	})
+
+	proxy := New(config)
+	defer proxy.StopProcesses()
+	req := httptest.NewRequest("GET", "/upstream/model1/test", nil)
+	rec := httptest.NewRecorder()
+	proxy.HandlerFunc(rec, req)
+	assert.Equal(t, http.StatusOK, rec.Code)
+	assert.Equal(t, "model1", rec.Body.String())
 }

proxy/sanitize_cors.go

@@ -0,0 +1,43 @@
+package proxy
+
+import (
+	"strings"
+)
+
+func isTokenChar(r rune) bool {
+	switch {
+	case r >= 'a' && r <= 'z':
+	case r >= 'A' && r <= 'Z':
+	case r >= '0' && r <= '9':
+	case strings.ContainsRune("!#$%&'*+-.^_`|~", r):
+	default:
+		return false
+	}
+	return true
+}
+
+func SanitizeAccessControlRequestHeaderValues(headerValues string) string {
+	parts := strings.Split(headerValues, ",")
+	valid := make([]string, 0, len(parts))
+
+	for _, p := range parts {
+		v := strings.TrimSpace(p)
+		if v == "" {
+			continue
+		}
+
+		validPart := true
+		for _, c := range v {
+			if !isTokenChar(c) {
+				validPart = false
+				break
+			}
+		}
+
+		if validPart {
+			valid = append(valid, v)
+		}
+	}
+
+	return strings.Join(valid, ", ")
+}

proxy/sanitize_cors_test.go

@@ -0,0 +1,77 @@
+package proxy
+
+import "testing"
+
+func TestSanitizeAccessControlRequestHeaderValues(t *testing.T) {
+	tests := []struct {
+		name     string
+		input    string
+		expected string
+	}{
+		{
+			name:     "empty string",
+			input:    "",
+			expected: "",
+		},
+		{
+			name:     "whitespace only",
+			input:    "   ",
+			expected: "",
+		},
+		{
+			name:     "single valid value",
+			input:    "content-type",
+			expected: "content-type",
+		},
+		{
+			name:     "multiple valid values",
+			input:    "content-type, authorization, x-requested-with",
+			expected: "content-type, authorization, x-requested-with",
+		},
+		{
+			name:     "values with extra spaces",
+			input:    "  content-type  ,  authorization  ",
+			expected: "content-type, authorization",
+		},
+		{
+			name:     "values with tabs",
+			input:    "content-type,\tauthorization",
+			expected: "content-type, authorization",
+		},
+		{
+			name:     "values with invalid characters",
+			input:    "content-type, auth\n, x-requested-with\r",
+			expected: "content-type, auth, x-requested-with",
+		},
+		{
+			name:     "empty values in list",
+			input:    "content-type,,authorization",
+			expected: "content-type, authorization",
+		},
+		{
+			name:     "leading and trailing commas",
+			input:    ",content-type,authorization,",
+			expected: "content-type, authorization",
+		},
+		{
+			name:     "mixed valid and invalid values",
+			input:    "content-type, \x00invalid, x-requested-with",
+			expected: "content-type, x-requested-with",
+		},
+		{
+			name:     "mixed case values",
+			input:    "Content-Type, my-Valid-Header, Another-hEader",
+			expected: "Content-Type, my-Valid-Header, Another-hEader",
+		},
+	}
+
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			got := SanitizeAccessControlRequestHeaderValues(tt.input)
+			if got != tt.expected {
+				t.Errorf("SanitizeAccessControlRequestHeaderValues(%q) = %q, want %q",
+					tt.input, got, tt.expected)
+			}
+		})
+	}
+}