feat: missing defaults added and readme additions

README.md

@@ -115,10 +115,8 @@ For additional configurations in HTTP server setup, please read [this](https://g
 
    func (baseAccount *BaseAccount) GetConfigForProvider(providerKey schemas.ModelProvider) (*schemas.ProviderConfig, error) {
        return &schemas.ProviderConfig{
-         ConcurrencyAndBufferSize: schemas.ConcurrencyAndBufferSize{
-           Concurrency: 3,
-           BufferSize:  10,
-         },
+          NetworkConfig:            schemas.DefaultNetworkConfig,
+          ConcurrencyAndBufferSize: schemas.DefaultConcurrencyAndBufferSize,
        }, nil
    }
    ```
@@ -246,15 +244,17 @@ Bifrost is divided into three Go packages: core, plugins, and transports.
 2. **plugins**: This package serves as an extension to core. You can download individual packages using `go get github.com/maximhq/bifrost/plugins/{plugin-name}` and pass the plugins while initializing Bifrost.
 
 ```golang
-plugin, err := plugins.NewMaximLoggerPlugin(os.Getenv("MAXIM_API_KEY"), os.Getenv("MAXIM_LOGGER_ID"))
+// go get github.com/maximhq/bifrost/plugins/maxim
+
+maximPlugin, err := maxim.NewMaximLoggerPlugin(os.Getenv("MAXIM_API_KEY"), os.Getenv("MAXIM_LOGGER_ID"))
 if err != nil {
   return nil, err
 }
 
 // Initialize Bifrost
 client, err := bifrost.Init(schemas.BifrostConfig{
   Account: &account,
-  Plugins: []schemas.Plugin{plugin},
+  Plugins: []schemas.Plugin{maximPlugin},
 })
 ```
 

core/schemas/provider.go

@@ -30,6 +30,14 @@ type NetworkConfig struct {
 	RetryBackoffMax                time.Duration `json:"retry_backoff_max"`                  // Maximum backoff duration
 }
 
+// DefaultNetworkConfig is the default network configuration for provider connections.
+var DefaultNetworkConfig = NetworkConfig{
+	DefaultRequestTimeoutInSeconds: DefaultRequestTimeoutInSeconds,
+	MaxRetries:                     DefaultMaxRetries,
+	RetryBackoffInitial:            DefaultRetryBackoffInitial,
+	RetryBackoffMax:                DefaultRetryBackoffMax,
+}
+
 // MetaConfig defines the interface for provider-specific configuration.
 // Check /meta folder for implemented provider-specific meta configurations.
 type MetaConfig interface {
@@ -61,6 +69,12 @@ type ConcurrencyAndBufferSize struct {
 	BufferSize  int `json:"buffer_size"` // Size of the buffer
 }
 
+// DefaultConcurrencyAndBufferSize is the default concurrency and buffer size for provider operations.
+var DefaultConcurrencyAndBufferSize = ConcurrencyAndBufferSize{
+	Concurrency: DefaultConcurrency,
+	BufferSize:  DefaultBufferSize,
+}
+
 // ProxyType defines the type of proxy to use for connections.
 type ProxyType string
 

core/tests/account.go

@@ -136,16 +136,8 @@ func (baseAccount *BaseAccount) GetConfigForProvider(providerKey schemas.ModelPr
 		}, nil
 	case schemas.Anthropic:
 		return &schemas.ProviderConfig{
-			NetworkConfig: schemas.NetworkConfig{
-				DefaultRequestTimeoutInSeconds: 30,
-				MaxRetries:                     1,
-				RetryBackoffInitial:            100 * time.Millisecond,
-				RetryBackoffMax:                2 * time.Second,
-			},
-			ConcurrencyAndBufferSize: schemas.ConcurrencyAndBufferSize{
-				Concurrency: 3,
-				BufferSize:  10,
-			},
+			NetworkConfig:            schemas.DefaultNetworkConfig,
+			ConcurrencyAndBufferSize: schemas.DefaultConcurrencyAndBufferSize,
 		}, nil
 	case schemas.Bedrock:
 		return &schemas.ProviderConfig{
@@ -166,16 +158,8 @@ func (baseAccount *BaseAccount) GetConfigForProvider(providerKey schemas.ModelPr
 		}, nil
 	case schemas.Cohere:
 		return &schemas.ProviderConfig{
-			NetworkConfig: schemas.NetworkConfig{
-				DefaultRequestTimeoutInSeconds: 30,
-				MaxRetries:                     1,
-				RetryBackoffInitial:            100 * time.Millisecond,
-				RetryBackoffMax:                2 * time.Second,
-			},
-			ConcurrencyAndBufferSize: schemas.ConcurrencyAndBufferSize{
-				Concurrency: 3,
-				BufferSize:  10,
-			},
+			NetworkConfig:            schemas.DefaultNetworkConfig,
+			ConcurrencyAndBufferSize: schemas.DefaultConcurrencyAndBufferSize,
 		}, nil
 	case schemas.Azure:
 		return &schemas.ProviderConfig{

plugins/maxim/README.md

@@ -0,0 +1,85 @@
+# Maxim-SDK Plugin for Bifrost
+
+This plugin integrates the Maxim SDK into Bifrost, enabling seamless observability and evaluation of LLM interactions. It captures and forwards inputs/outputs from Bifrost to the Maxim's observability platform. This facilitates end-to-end tracing, evaluation, and monitoring of your LLM-based application.
+
+## Usage for Bifrost Go Package
+
+1. Download the Plugin
+
+```bash
+go get github.com/maximhq/bifrost/plugins/maxim
+```
+
+2. Initialise the Plugin
+
+```go
+    maximPlugin, err := maxim.NewMaximLoggerPlugin("your_maxim_api_key", "your_maxim_log_repo_id")
+    if err != nil {
+        return nil, err
+        }
+```
+
+3.  Pass to plugin to bifrost
+
+```go
+    client, err := bifrost.Init(schemas.BifrostConfig{
+        Account: &yourAccount,
+        Plugins: []schemas.Plugin{maximPlugin},
+        })
+```
+
+## Usage for Bifrost HTTP Transport
+
+1. Set up the environment variables
+
+```bash
+export MAXIM_API_KEY=your_maxim_api_key
+```
+
+2. Setup flags to use the plugin
+   Use include `maxim` in `--plugins` and your maxim log repo id in `--maxim-log-repo-id"`
+
+   eg. `bifrost-http -config config.json -env .env -plugins maxim -maxim-log-repo-id your_maxim_log_repo_id`
+
+   For docker build
+
+   ```bash
+   docker build \
+   --build-arg CONFIG_PATH=./config.example.json \
+   --build-arg ENV_PATH=./.env.sample \
+   --build-arg PORT=8080 \
+   --build-arg POOL_SIZE=300 \
+   --build-arg PLUGINS=maxim \
+   --build-arg MAXIM_LOG_REPO_ID=your_maxim_log_repo_id \
+   -t bifrost-transports .
+   ```
+
+## Viewing Your Traces
+
+1. Log in to your [Maxim Dashboard](https://getmaxim.ai/dashboard)
+2. Navigate to your repository
+3. View detailed llm traces, including:
+   - LLM inputs/outputs
+   - Tool usage patterns
+   - Performance metrics
+   - Cost analytics
+
+## Additional Features
+
+The plugin also supports custom `trace-id` and `generation-id` if the uses wish to log the generations to their custom logging implementation. To use it, just pass your trace id to the passed request context with the key `trace-id`, and similarly to `generation-id` for generation id. In these cases no new trace/generation is created and the output is just logged to your provided generation.
+
+eg.
+
+```go
+    ctx = context.WithValue(ctx, "generation-id", "123")
+
+    result, err := bifrostClient.ChatCompletionRequest(schemas.OpenAI, &schemas.BifrostRequest{
+        Model: "gpt-4o",
+        Input: schemas.RequestInput{
+            ChatCompletionInput: &messages,
+            },
+            Params: &params,
+            }, ctx)
+```
+
+HTTP transport offers out of the box support for this feature(when maxim plugin is used), just pass `x-bf-maxim-trace-id` of `x-bf-maxim-generation-id` header with your request to use this feature.

transports/Dockerfile

@@ -38,13 +38,15 @@ ARG POOL_SIZE
 ARG DROP_EXCESS_REQUESTS
 ARG PLUGINS
 ARG MAXIM_LOG_REPO_ID
+ARG PROMETHEUS_LABELS
 
 # Set default values if args are not provided
 ENV APP_PORT=${PORT:-8080}
 ENV APP_POOL_SIZE=${POOL_SIZE:-300}
 ENV APP_DROP_EXCESS_REQUESTS=${DROP_EXCESS_REQUESTS:-false}
 ENV APP_PLUGINS=${PLUGINS:-""}
 ENV APP_MAXIM_LOG_REPO_ID=${MAXIM_LOG_REPO_ID:-""}
+ENV APP_PROMETHEUS_LABELS=${PROMETHEUS_LABELS:-""}
 
 # Copy the config and environment files into the image
 COPY ${CONFIG_PATH} /app/config/config.json
@@ -55,7 +57,7 @@ RUN echo '#!/bin/sh' > /app/entrypoint.sh && \
     echo 'if [ ! -f /app/config/config.json ]; then echo "Missing config.json"; exit 1; fi' >> /app/entrypoint.sh && \
     echo 'if [ ! -f /app/config/.env ]; then echo "Missing .env"; exit 1; fi' >> /app/entrypoint.sh && \
     echo 'if [ ! -f /app/main ]; then echo "Missing main binary"; exit 1; fi' >> /app/entrypoint.sh && \
-    echo 'exec /app/main -config /app/config/config.json -env /app/config/.env -port "$APP_PORT" -pool-size "$APP_POOL_SIZE" -drop-excess-requests "$APP_DROP_EXCESS_REQUESTS" -plugins "$APP_PLUGINS" -maxim-log-repo-id "$APP_MAXIM_LOG_REPO_ID"' >> /app/entrypoint.sh && \
+    echo 'exec /app/main -config /app/config/config.json -env /app/config/.env -port "$APP_PORT" -pool-size "$APP_POOL_SIZE" -drop-excess-requests "$APP_DROP_EXCESS_REQUESTS" -plugins "$APP_PLUGINS" -maxim-log-repo-id "$APP_MAXIM_LOG_REPO_ID" -prometheus-labels "$APP_PROMETHEUS_LABELS"' >> /app/entrypoint.sh && \
     chmod +x /app/entrypoint.sh
 
 # Expose the port defined by argument

transports/README.md

@@ -15,6 +15,8 @@ This package contains clients for various transports that can be used to spin up
     - [Text Completions](#text-completions)
     - [Chat Completions](#chat-completions)
   - [🔧 Advanced Features](#-advanced-features)
+    - [Prometheus Support](#prometheus-support)
+    - [Plugin Support](#plugin-support)
     - [Fallbacks](#fallbacks)
 
 ---
@@ -165,6 +167,22 @@ curl -X POST http://localhost:8080/v1/chat/completions \
 
 ## 🔧 Advanced Features
 
+### Prometheus Support
+
+HTTP transport supports Prometheus out of the box. By default all the metrics are available at `/metrics` endpoint. It providers metrics for httpRequestsTotal, httpRequestDuration, httpRequestSizeBytes, httpResponseSizeBytes, bifrostUpstreamRequestsTotal, and bifrostUpstreamLatencySeconds. To add custom labels to these metrics using can pass a flag of `-prometheus-labels` while running the http transport.
+
+For eg. `-prometheus-labels team-id,task-id,location`
+
+Values for labels are then picked up from the HTTP request headers with the prefix `x-bf-prom-`.
+
+### Plugin Support
+
+You can explore the available plugins [here](https://github.com/maximhq/bifrost/tree/main/plugins). And to attached these plugins to your HTTP transport, just pass the flag `-plugins`.
+
+For eg. `-plugins maxim`
+
+Note: Please check plugin specific documentations (github.com/maximhq/bifrost/tree/main/plugins/{plugin_name}) for more nuanced control.
+
 ### Fallbacks
 
 Configure fallback options in your requests: