diff --git a/Argcfile.sh b/Argcfile.sh index 96300b4..5f04b6f 100755 --- a/Argcfile.sh +++ b/Argcfile.sh @@ -229,27 +229,6 @@ models-cohere() { } -# @cmd Chat with reka api -# @env REKA_API_KEY! -# @option -m --model=reka-flash $REKA_MODEL -# @flag -S --no-stream -# @arg text~ -chat-reka() { - _wrapper curl -i https://api.reka.ai/v1/chat \ --X POST \ --H 'Content-Type: application/json' \ --H "X-Api-Key: $REKA_API_KEY" \ --d "$(_build_body reka "$@")" -} - -# @cmd List reka models -# @env REKA_API_KEY! -models-reka() { - _wrapper curl https://api.reka.ai/v1/models \ --H "X-Api-Key: $REKA_API_KEY" \ - -} - # @cmd Chat with ollama api # @option -m --model=codegemma $OLLAMA_MODEL # @flag -S --no-stream @@ -492,18 +471,6 @@ _build_body() { "model": "'$argc_model'", "message": "'"$*"'", "stream": '$stream' -}' - ;; - reka) - echo '{ - "model": "'$argc_model'", - "messages": [ - { - "role": "user", - "content": "'"$*"'" - } - ], - "stream": '$stream' }' ;; claude) diff --git a/README.md b/README.md index faf21dd..69c73a9 100644 --- a/README.md +++ b/README.md @@ -31,7 +31,6 @@ AIChat is an all-in-one AI CLI tool featuring Chat-REPL, RAG, Function Calling, - Claude: Claude-3.5/Claude-3 (paid, vision, function-calling) - Mistral (paid, embedding, function-calling) - Cohere: Command-R/Command-R+ (paid, embedding, rerank, function-calling) -- Reka (paid, vision) - Perplexity: Llama-3/Mixtral (paid) - Groq: Llama-3/Mixtral/Gemma (free, function-calling) - Ollama (free, local, embedding) diff --git a/config.example.yaml b/config.example.yaml index 8425cc7..318db2d 100644 --- a/config.example.yaml +++ b/config.example.yaml @@ -144,10 +144,6 @@ clients: - type: cohere api_key: xxx # ENV: {client}_API_KEY - # See https://docs.reka.ai/quick-start - - type: reka - api_key: xxx # ENV: {client}_API_KEY - # See https://docs.perplexity.ai/docs/getting-started - type: openai-compatible name: perplexity diff --git a/models.yaml b/models.yaml index 246c607..0cfe544 100644 --- a/models.yaml +++ b/models.yaml @@ -189,28 +189,6 @@ type: rerank max_input_tokens: 4096 -- platform: reka - docs: - # - https://www.reka.ai/ourmodels - # - https://www.reka.ai/reka-deploy - # - https://docs.reka.ai/api-reference/chat/create - models: - - name: reka-core - max_input_tokens: 128000 - input_price: 10 - output_price: 25 - supports_vision: true - - name: reka-flash - max_input_tokens: 128000 - input_price: 0.8 - output_price: 2 - supports_vision: true - - name: reka-edge - max_input_tokens: 128000 - input_price: 0.4 - output_price: 1 - supports_vision: true - - platform: perplexity # docs: # - https://docs.perplexity.ai/docs/model-cards diff --git a/src/client/mod.rs b/src/client/mod.rs index 6710e20..579c2a3 100644 --- a/src/client/mod.rs +++ b/src/client/mod.rs @@ -24,7 +24,6 @@ register_client!( (gemini, "gemini", GeminiConfig, GeminiClient), (claude, "claude", ClaudeConfig, ClaudeClient), (cohere, "cohere", CohereConfig, CohereClient), - (reka, "reka", RekaConfig, RekaClient), (ollama, "ollama", OllamaConfig, OllamaClient), ( azure_openai, diff --git a/src/client/reka.rs b/src/client/reka.rs deleted file mode 100644 index 2e9b88f..0000000 --- a/src/client/reka.rs +++ /dev/null @@ -1,124 +0,0 @@ -use super::*; - -use anyhow::{anyhow, Result}; -use reqwest::{Client as ReqwestClient, RequestBuilder}; -use serde::Deserialize; -use serde_json::{json, Value}; - -const CHAT_COMPLETIONS_API_URL: &str = "https://api.reka.ai/v1/chat"; - -#[derive(Debug, Clone, Deserialize, Default)] -pub struct RekaConfig { - pub name: Option, - pub api_key: Option, - #[serde(default)] - pub models: Vec, - pub patches: Option, - pub extra: Option, -} - -impl RekaClient { - config_get_fn!(api_key, get_api_key); - - pub const PROMPTS: [PromptAction<'static>; 1] = - [("api_key", "API Key:", true, PromptKind::String)]; - - fn chat_completions_builder( - &self, - client: &ReqwestClient, - data: ChatCompletionsData, - ) -> Result { - let api_key = self.get_api_key()?; - - let mut body = build_chat_completions_body(data, &self.model); - self.patch_chat_completions_body(&mut body); - - let url = CHAT_COMPLETIONS_API_URL; - - debug!("Reka Chat Completions Request: {url} {body}"); - - let builder = client.post(url).header("x-api-key", api_key).json(&body); - - Ok(builder) - } -} - -impl_client_trait!(RekaClient, chat_completions, chat_completions_streaming); - -async fn chat_completions(builder: RequestBuilder) -> Result { - let res = builder.send().await?; - let status = res.status(); - let data: Value = res.json().await?; - if !status.is_success() { - catch_error(&data, status.as_u16())?; - } - - debug!("non-stream-data: {data}"); - extract_chat_completions(&data) -} - -async fn chat_completions_streaming( - builder: RequestBuilder, - handler: &mut SseHandler, -) -> Result<()> { - let mut prev_text = String::new(); - let handle = |message: SseMmessage| -> Result { - let data: Value = serde_json::from_str(&message.data)?; - debug!("stream-data: {data}"); - if let Some(text) = data["responses"][0]["chunk"]["content"].as_str() { - let delta_text = &text[prev_text.len()..]; - prev_text = text.to_string(); - handler.text(delta_text)?; - } - Ok(false) - }; - - sse_stream(builder, handle).await -} - -fn build_chat_completions_body(data: ChatCompletionsData, model: &Model) -> Value { - let ChatCompletionsData { - mut messages, - temperature, - top_p, - functions: _, - stream, - } = data; - - patch_system_message(&mut messages); - - let mut body = json!({ - "model": &model.name(), - "messages": messages, - }); - - if let Some(v) = model.max_tokens_param() { - body["max_tokens"] = v.into(); - } - if let Some(v) = temperature { - body["temperature"] = v.into(); - } - if let Some(v) = top_p { - body["top_p"] = v.into(); - } - if stream { - body["stream"] = true.into(); - } - - body -} - -fn extract_chat_completions(data: &Value) -> Result { - let text = data["responses"][0]["message"]["content"] - .as_str() - .ok_or_else(|| anyhow!("Invalid response data: {data}"))?; - - let output = ChatCompletionsOutput { - text: text.to_string(), - tool_calls: vec![], - id: data["id"].as_str().map(|v| v.to_string()), - input_tokens: data["usage"]["input_tokens"].as_u64(), - output_tokens: data["usage"]["output_tokens"].as_u64(), - }; - Ok(output) -}