💄 style: update Together AI model list (lobehub#3713)

* 💄 style: update Together AI model list * 💄 style: update model info * 💄 style: update Llama3.1 405B tokens num * 🔨 chore: fix CI
qinyuhang · Sep 1, 2024 · 0dde3b1 · 0dde3b1
1 parent 2ae1ba2
commit 0dde3b1
Show file tree

Hide file tree

Showing 2 changed files with 123 additions and 38 deletions.
diff --git a/src/config/modelProviders/togetherai.ts b/src/config/modelProviders/togetherai.ts
@@ -1,76 +1,161 @@
 import { ModelProviderCard } from '@/types/llm';
 
-// ref https://docs.together.ai/docs/inference-models
+// ref https://docs.together.ai/docs/chat-models
+// ref https://www.together.ai/pricing
 const TogetherAI: ModelProviderCard = {
   chatModels: [
     {
-      displayName: 'Deepseek Coder Instruct (33B)',
+      displayName: 'Llama 3.1 8B Instruct Turbo',
       enabled: true,
-      id: 'deepseek-ai/deepseek-coder-33b-instruct',
-      tokens: 16_384,
+      id: 'meta-llama/Meta-Llama-3.1-8B-Instruct-Turbo',
+      tokens: 131_072,
     },
     {
-      displayName: 'Phind Code LLaMA v2 (34B)',
+      displayName: 'Llama 3.1 70B Instruct Turbo',
       enabled: true,
-      id: 'Phind/Phind-CodeLlama-34B-v2',
-      tokens: 16_384,
+      id: 'meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo',
+      tokens: 131_072,
     },
     {
-      displayName: 'Gemma Instruct (2B)',
+      displayName: 'Llama 3.1 405B Instruct Turbo',
       enabled: true,
-      id: 'google/gemma-2b-it',
+      id: 'meta-llama/Meta-Llama-3.1-405B-Instruct-Turbo',
+      tokens: 8192,
+    },
+    {
+      displayName: 'Llama 3 8B Instruct Turbo',
+      id: 'meta-llama/Meta-Llama-3-8B-Instruct-Turbo',
+      tokens: 8192,
+    },
+    {
+      displayName: 'Llama 3 70B Instruct Turbo',
+      id: 'meta-llama/Meta-Llama-3-70B-Instruct-Turbo',
+      tokens: 8192,
+    },
+    {
+      displayName: 'Llama 3 8B Instruct Lite',
+      id: 'meta-llama/Meta-Llama-3-8B-Instruct-Lite',
+      tokens: 8192,
+    },
+    {
+      displayName: 'Llama 3 70B Instruct Lite',
+      id: 'meta-llama/Meta-Llama-3-70B-Instruct-Lite',
+      tokens: 8192,
+    },
+    {
+      displayName: 'LLaMA-3 Chat (8B)',
+      id: 'meta-llama/Llama-3-8b-chat-hf',
+      tokens: 8192,
+    },
+    {
+      displayName: 'LLaMA-3 Chat (70B)',
+      id: 'meta-llama/Llama-3-70b-chat-hf',
       tokens: 8192,
     },
     {
       displayName: 'LLaMA-2 Chat (13B)',
-      enabled: true,
       id: 'meta-llama/Llama-2-13b-chat-hf',
       tokens: 4096,
     },
     {
-      displayName: '01-ai Yi Chat (34B)',
+      displayName: 'Gemma 2 9B',
       enabled: true,
-      id: 'zero-one-ai/Yi-34B-Chat',
-      tokens: 4096,
+      id: 'google/gemma-2-9b-it',
+      tokens: 8192,
     },
     {
-      displayName: 'Mixtral-8x7B Instruct (46.7B)',
+      displayName: 'Gemma 2 27B',
       enabled: true,
-      id: 'mistralai/Mixtral-8x7B-Instruct-v0.1',
+      id: 'google/gemma-2-27b-it',
+      tokens: 8192,
+    },
+    {
+      displayName: 'Gemma Instruct (2B)',
+      id: 'google/gemma-2b-it',
+      tokens: 8192,
+    },
+    {
+      displayName: 'Mistral (7B) Instruct v0.3',
+      enabled: true,
+      id: 'mistralai/Mistral-7B-Instruct-v0.3',
       tokens: 32_768,
     },
     {
-      displayName: 'Nous Hermes 2 - Mixtral 8x7B-DPO (46.7B)',
+      displayName: 'Mistral (7B) Instruct v0.2',
+      id: 'mistralai/Mistral-7B-Instruct-v0.2',
+      tokens: 32_768,
+    },
+    {
+      displayName: 'Mistral (7B) Instruct',
+      id: 'mistralai/Mistral-7B-Instruct-v0.1',
+      tokens: 8192,
+    },
+    {
+      displayName: 'Mixtral-8x7B Instruct (46.7B)',
       enabled: true,
-      id: 'NousResearch/Nous-Hermes-2-Mixtral-8x7B-DPO',
+      id: 'mistralai/Mixtral-8x7B-Instruct-v0.1',
       tokens: 32_768,
     },
     {
-      displayName: 'Nous Hermes-2 Yi (34B)',
+      displayName: 'Mixtral-8x22B Instruct (141B)',
       enabled: true,
-      id: 'NousResearch/Nous-Hermes-2-Yi-34B',
-      tokens: 4096,
+      id: 'mistralai/Mixtral-8x22B-Instruct-v0.1',
+      tokens: 65_536,
     },
     {
-      displayName: 'Qwen 1.5 Chat (7B)',
+      displayName: 'DeepSeek LLM Chat (67B)',
       enabled: true,
-      id: 'Qwen/Qwen1.5-7B-Chat',
-      tokens: 32_768,
+      id: 'deepseek-ai/deepseek-llm-67b-chat',
+      tokens: 4096,
     },
     {
-      displayName: 'Qwen 1.5 Chat (32B)',
+      displayName: 'Qwen 2 Instruct (72B)',
       enabled: true,
-      id: 'Qwen/Qwen1.5-32B-Chat',
+      id: 'Qwen/Qwen2-72B-Instruct',
       tokens: 32_768,
     },
     {
       displayName: 'Qwen 1.5 Chat (72B)',
-      enabled: true,
       id: 'Qwen/Qwen1.5-72B-Chat',
       tokens: 32_768,
     },
+    {
+      displayName: 'Qwen 1.5 Chat (110B)',
+      id: 'Qwen/Qwen1.5-110B-Chat',
+      tokens: 32_768,
+    },
+    {
+      displayName: 'DBRX Instruct',
+      id: 'databricks/dbrx-instruct',
+      tokens: 32_768,
+    },
+    {
+      displayName: 'Upstage SOLAR Instruct v1 (11B)',
+      id: 'upstage/SOLAR-10.7B-Instruct-v1.0',
+      tokens: 4096,
+    },
+    {
+      displayName: 'Nous Hermes 2 - Mixtral 8x7B-DPO (46.7B)',
+      id: 'NousResearch/Nous-Hermes-2-Mixtral-8x7B-DPO',
+      tokens: 32_768,
+    },
+    {
+      displayName: 'Nous Hermes-2 Yi (34B)',
+      id: 'NousResearch/Nous-Hermes-2-Yi-34B',
+      tokens: 4096,
+    },
+    {
+      displayName: 'MythoMax-L2 (13B)',
+      id: 'Gryphe/MythoMax-L2-13b',
+      tokens: 4096,
+    },
+    {
+      displayName: 'StripedHyena Nous (7B)',
+      id: 'togethercomputer/StripedHyena-Nous-7B',
+      tokens: 32_768,
+    },
   ],
-  checkModel: 'togethercomputer/alpaca-7b',
+  checkModel: 'meta-llama/Meta-Llama-3.1-8B-Instruct-Turbo',
   id: 'togetherai',
   modelList: { showModelFetcher: true },
   name: 'Together AI',

diff --git a/src/libs/agent-runtime/togetherai/__snapshots__/index.test.ts.snap b/src/libs/agent-runtime/togetherai/__snapshots__/index.test.ts.snap
@@ -45,7 +45,7 @@ exports[`LobeTogetherAI > models > should get models 1`] = `
   {
     "description": "Nous Hermes 2 Mixtral 7bx8 DPO is the new flagship Nous Research model trained over the Mixtral 7bx8 MoE LLM. The model was trained on over 1,000,000 entries of primarily GPT-4 generated data, as well as other high quality data from open datasets across the AI landscape, achieving state of the art performance on a variety of tasks.",
     "displayName": "Nous Hermes 2 - Mixtral 8x7B-DPO ",
-    "enabled": true,
+    "enabled": false,
     "functionCall": false,
     "id": "NousResearch/Nous-Hermes-2-Mixtral-8x7B-DPO",
     "maxOutput": 32768,
@@ -65,7 +65,7 @@ exports[`LobeTogetherAI > models > should get models 1`] = `
   {
     "description": "Nous Hermes 2 - Yi-34B is a state of the art Yi Fine-tune",
     "displayName": "Nous Hermes-2 Yi (34B)",
-    "enabled": true,
+    "enabled": false,
     "functionCall": false,
     "id": "NousResearch/Nous-Hermes-2-Yi-34B",
     "maxOutput": 4096,
@@ -145,7 +145,7 @@ exports[`LobeTogetherAI > models > should get models 1`] = `
   {
     "description": "Qwen1.5 is the beta version of Qwen2, a transformer-based decoder-only language model pretrained on a large amount of data. In comparison with the previous released Qwen.",
     "displayName": "Qwen 1.5 Chat (32B)",
-    "enabled": true,
+    "enabled": false,
     "functionCall": false,
     "id": "Qwen/Qwen1.5-32B-Chat",
     "maxOutput": 32768,
@@ -165,7 +165,7 @@ exports[`LobeTogetherAI > models > should get models 1`] = `
   {
     "description": "Qwen1.5 is the beta version of Qwen2, a transformer-based decoder-only language model pretrained on a large amount of data. In comparison with the previous released Qwen.",
     "displayName": "Qwen 1.5 Chat (72B)",
-    "enabled": true,
+    "enabled": false,
     "functionCall": false,
     "id": "Qwen/Qwen1.5-72B-Chat",
     "maxOutput": 32768,
@@ -175,7 +175,7 @@ exports[`LobeTogetherAI > models > should get models 1`] = `
   {
     "description": "Qwen1.5 is the beta version of Qwen2, a transformer-based decoder-only language model pretrained on a large amount of data. In comparison with the previous released Qwen.",
     "displayName": "Qwen 1.5 Chat (7B)",
-    "enabled": true,
+    "enabled": false,
     "functionCall": false,
     "id": "Qwen/Qwen1.5-7B-Chat",
     "maxOutput": 32768,
@@ -295,7 +295,7 @@ exports[`LobeTogetherAI > models > should get models 1`] = `
   {
     "description": "Deepseek Coder is composed of a series of code language models, each trained from scratch on 2T tokens, with a composition of 87% code and 13% natural language in both English and Chinese.",
     "displayName": "Deepseek Coder Instruct (33B)",
-    "enabled": true,
+    "enabled": false,
     "functionCall": false,
     "id": "deepseek-ai/deepseek-coder-33b-instruct",
     "maxOutput": 16384,
@@ -305,7 +305,7 @@ exports[`LobeTogetherAI > models > should get models 1`] = `
   {
     "description": "trained from scratch on a vast dataset of 2 trillion tokens in both English and Chinese",
     "displayName": "DeepSeek LLM Chat (67B)",
-    "enabled": false,
+    "enabled": true,
     "functionCall": false,
     "id": "deepseek-ai/deepseek-llm-67b-chat",
     "maxOutput": 4096,
@@ -325,7 +325,7 @@ exports[`LobeTogetherAI > models > should get models 1`] = `
   {
     "description": "Gemma is a family of lightweight, state-of-the-art open models from Google, built from the same research and technology used to create the Gemini models.",
     "displayName": "Gemma Instruct (2B)",
-    "enabled": true,
+    "enabled": false,
     "functionCall": false,
     "id": "google/gemma-2b-it",
     "maxOutput": 8192,
@@ -365,7 +365,7 @@ exports[`LobeTogetherAI > models > should get models 1`] = `
   {
     "description": "Llama 2-chat leverages publicly available instruction datasets and over 1 million human annotations. Available in three sizes: 7B, 13B and 70B parameters",
     "displayName": "LLaMA-2 Chat (13B)",
-    "enabled": true,
+    "enabled": false,
     "functionCall": false,
     "id": "meta-llama/Llama-2-13b-chat-hf",
     "maxOutput": 4096,
@@ -445,7 +445,7 @@ exports[`LobeTogetherAI > models > should get models 1`] = `
   {
     "description": "The Mixtral-8x22B-Instruct-v0.1 Large Language Model (LLM) is an instruct fine-tuned version of the Mixtral-8x22B-v0.1.",
     "displayName": "Mixtral-8x22B Instruct v0.1",
-    "enabled": false,
+    "enabled": true,
     "functionCall": false,
     "id": "mistralai/Mixtral-8x22B-Instruct-v0.1",
     "maxOutput": 65536,
@@ -565,7 +565,7 @@ exports[`LobeTogetherAI > models > should get models 1`] = `
   {
     "description": "The Yi series models are large language models trained from scratch by developers at 01.AI",
     "displayName": "01-ai Yi Chat (34B)",
-    "enabled": true,
+    "enabled": false,
     "functionCall": false,
     "id": "zero-one-ai/Yi-34B-Chat",
     "maxOutput": 4096,