fix: update model descriptions in catalog (#889)

Fixes #887

Signed-off-by: Jeff MAURY <[email protected]>

Author: jeffmaury

packages/backend/src/assets/ai.json

@@ -79,7 +79,7 @@
     {
       "id": "hf.TheBloke.mistral-7b-instruct-v0.1.Q4_K_M",
       "name": "TheBloke/Mistral-7B-Instruct-v0.1-GGUF",
-      "description": "The Mistral-7B-Instruct-v0.1 Large Language Model (LLM) is a instruct fine-tuned version of the [Mistral-7B-v0.1](https://huggingface.co/mistralai/Mistral-7B-v0.1) generative text model using a variety of publicly available conversation datasets. For full details of this model please read our [release blog post](https://mistral.ai/news/announcing-mistral-7b/)",
+      "description": "# Mistral 7B Instruct v0.1 - GGUF\n- Model creator: [Mistral AI](https://huggingface.co/mistralai)\n- Original model: [Mistral 7B Instruct v0.1](https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.1)\n\n<!-- description start -->\n## Description\n\nThis repo contains GGUF format model files for [Mistral AI's Mistral 7B Instruct v0.1](https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.1).\n",
       "hw": "CPU",
       "registry": "Hugging Face",
       "license": "Apache-2.0",
@@ -89,7 +89,7 @@
     {
       "id": "hf.NousResearch.Hermes-2-Pro-Mistral-7B.Q4_K_M",
       "name": "NousResearch/Hermes-2-Pro-Mistral-7B-GGUF",
-      "description": "This is the GGUF version of the model, made for the llama.cpp inference engine.\n If you are looking for the transformers/fp16 model, it is available here: [https://huggingface.co/NousResearch/Hermes-2-Pro-Mistral-7B](https://huggingface.co/NousResearch/Hermes-2-Pro-Mistral-7B)\n Hermes 2 Pro on Mistral 7B is the new flagship 7B Hermes!\n Hermes 2 Pro is an upgraded, retrained version of Nous Hermes 2, consisting of an updated and cleaned version of the OpenHermes 2.5 Dataset, as well as a newly introduced Function Calling and JSON Mode dataset developed in-house.\n This new version of Hermes maintains its excellent general task and conversation capabilities - but also excels at Function Calling, JSON Structured Outputs, and has improved on several other metrics as well, scoring a 90% on our function calling evaluation built in partnership with Fireworks.AI, and an 81% on our structured JSON Output evaluation.\n Hermes Pro takes advantage of a special system prompt and multi-turn function calling structure with a new chatml role in order to make function calling reliable and easy to parse. Learn more about prompting below.\nThis work was a collaboration between Nous Research, @interstellarninja, and Fireworks.AI\n Learn more about the function calling on our github repo here: [https://github.com/NousResearch/Hermes-Function-Calling/tree/main](https://github.com/NousResearch/Hermes-Function-Calling/tree/main)",
+      "description": "## Model Description\n\nHermes 2 Pro on Mistral 7B is the new flagship 7B Hermes!\n\nHermes 2 Pro is an upgraded, retrained version of Nous Hermes 2, consisting of an updated and cleaned version of the OpenHermes 2.5 Dataset, as well as a newly introduced Function Calling and JSON Mode dataset developed in-house.\n\nThis new version of Hermes maintains its excellent general task and conversation capabilities - but also excels at Function Calling, JSON Structured Outputs, and has improved on several other metrics as well, scoring a 90% on our function calling evaluation built in partnership with Fireworks.AI, and an 84% on our structured JSON Output evaluation.\n\nHermes Pro takes advantage of a special system prompt and multi-turn function calling structure with a new chatml role in order to make function calling reliable and easy to parse. Learn more about prompting below.\n\nThis work was a collaboration between Nous Research, @interstellarninja, and Fireworks.AI\n\nLearn more about the function calling system for this model on our github repo here: https://github.com/NousResearch/Hermes-Function-Calling",
       "hw": "CPU",
       "registry": "Hugging Face",
       "license": "Apache-2.0",
@@ -99,7 +99,7 @@
     {
       "id": "hf.ibm.merlinite-7b-Q4_K_M",
       "name": "ibm/merlinite-7b-GGUF",
-      "description": "## Merlinite 7b - GGUF\n4-bit quantized version of [ibm/merlinite-7b](https://huggingface.co/ibm/merlinite-7b)",
+      "description": "# Merlinite 7b - GGUF\n\n4-bit quantized version of [ibm/merlinite-7b](https://huggingface.co/ibm/merlinite-7b)",
       "hw": "CPU",
       "registry": "Hugging Face",
       "license": "Apache-2.0",
@@ -109,7 +109,7 @@
     {
       "id": "hf.TheBloke.mistral-7b-codealpaca-lora.Q4_K_M",
       "name": "TheBloke/Mistral-7B-codealpaca-lora-GGUF",
-      "description": "## Mistral 7B CodeAlpaca Lora - GGUF\n- Model creator: [Kamil](https://huggingface.co/Nondzu)\n- Original model: [Mistral 7B CodeAlpaca Lora](https://huggingface.co/Nondzu/Mistral-7B-codealpaca-lora)\n### Description\nThis repo contains GGUF format model files for [Kamil's Mistral 7B CodeAlpaca Lora](https://huggingface.co/Nondzu/Mistral-7B-codealpaca-lora).\nThese files were quantised using hardware kindly provided by [Massed Compute](https://massedcompute.com/).",
+      "description": "# Mistral 7B CodeAlpaca Lora - GGUF\n- Model creator: [Kamil](https://huggingface.co/Nondzu)\n- Original model: [Mistral 7B CodeAlpaca Lora](https://huggingface.co/Nondzu/Mistral-7B-codealpaca-lora)\n\n<!-- description start -->\n## Description\n\nThis repo contains GGUF format model files for [Kamil's Mistral 7B CodeAlpaca Lora](https://huggingface.co/Nondzu/Mistral-7B-codealpaca-lora).\n\nThese files were quantised using hardware kindly provided by [Massed Compute](https://massedcompute.com/).\n",
       "hw": "CPU",
       "registry": "Hugging Face",
       "license": "Apache-2.0",
@@ -119,7 +119,7 @@
     {
       "id": "hf.TheBloke.mistral-7b-code-16k-qlora.Q4_K_M",
       "name": "TheBloke/Mistral-7B-Code-16K-qlora-GGUF",
-      "description": "## Mistral 7B Code 16K qLoRA - GGUF\n- Model creator: [Kamil](https://huggingface.co/Nondzu)\n- Original model: [Mistral 7B Code 16K qLoRA](https://huggingface.co/Nondzu/Mistral-7B-code-16k-qlora)\n## Description\nThis repo contains GGUF format model files for [Kamil's Mistral 7B Code 16K qLoRA](https://huggingface.co/Nondzu/Mistral-7B-code-16k-qlora).",
+      "description": "# Mistral 7B Code 16K qLoRA - GGUF\n- Model creator: [Kamil](https://huggingface.co/Nondzu)\n- Original model: [Mistral 7B Code 16K qLoRA](https://huggingface.co/Nondzu/Mistral-7B-code-16k-qlora)\n\n<!-- description start -->\n## Description\n\nThis repo contains GGUF format model files for [Kamil's Mistral 7B Code 16K qLoRA](https://huggingface.co/Nondzu/Mistral-7B-code-16k-qlora).",
       "hw": "CPU",
       "registry": "Hugging Face",
       "license": "Apache-2.0",
@@ -129,7 +129,7 @@
     {
       "id": "hf.froggeric.Cerebrum-1.0-7b-Q4_KS",
       "name": "froggeric/Cerebrum-1.0-7b-GGUF",
-      "description": "GGUF quantisations of [AetherResearch/Cerebrum-1.0-7b](https://huggingface.co/AetherResearch/Cerebrum-1.0-7b)\n## Introduction\nCerebrum 7b is a large language model (LLM) created specifically for reasoning tasks. It is based on the Mistral 7b model, fine-tuned on a small custom dataset of native chain of thought data and further improved with targeted RLHF (tRLHF), a novel technique for sample-efficient LLM alignment. Unlike numerous other recent fine-tuning approaches, our training pipeline includes under 5000 training prompts and even fewer labeled datapoints for tRLHF.\nNative chain of thought approach means that Cerebrum is trained to devise a tactical plan before tackling problems that require thinking. For brainstorming, knowledge intensive, and creative tasks Cerebrum will typically omit unnecessarily verbose considerations.\nZero-shot prompted Cerebrum significantly outperforms few-shot prompted Mistral 7b as well as much larger models (such as Llama 2 70b) on a range of tasks that require reasoning, including ARC Challenge, GSM8k, and Math.\nThis LLM model works a lot better than any other mistral mixtral models for agent data, tested on 14th March 2024.",
+      "description": "GGUF quantisations of [AetherResearch/Cerebrum-1.0-7b](https://huggingface.co/AetherResearch/Cerebrum-1.0-7b)\n\n## Introduction\n\nCerebrum 7b is a large language model (LLM) created specifically for reasoning tasks. It is based on the Mistral 7b model, fine-tuned on a small custom dataset of native chain of thought data and further improved with targeted RLHF (tRLHF), a novel technique for sample-efficient LLM alignment. Unlike numerous other recent fine-tuning approaches, our training pipeline includes under 5000 training prompts and even fewer labeled datapoints for tRLHF.\n\nNative chain of thought approach means that Cerebrum is trained to devise a tactical plan before tackling problems that require thinking. For brainstorming, knowledge intensive, and creative tasks Cerebrum will typically omit unnecessarily verbose considerations.\n\nZero-shot prompted Cerebrum significantly outperforms few-shot prompted Mistral 7b as well as much larger models (such as Llama 2 70b) on a range of tasks that require reasoning, including ARC Challenge, GSM8k, and Math.\n\nThis LLM model works a lot better than any other mistral mixtral models for agent data, tested on 14th March 2024.\n",
       "hw": "CPU",
       "registry": "Hugging Face",
       "license": "Apache-2.0",
@@ -139,7 +139,7 @@
     {
       "id": "hf.TheBloke.openchat-3.5-0106.Q4_K_M",
       "name": "TheBloke/openchat-3.5-0106-GGUF",
-      "description": "## Openchat 3.5 0106 - GGUF\n- Model creator: [OpenChat](https://huggingface.co/openchat)\n- Original model: [Openchat 3.5 0106](https://huggingface.co/openchat/openchat-3.5-0106)\n## DescriptionThis repo contains GGUF format model files for [OpenChat's Openchat 3.5 0106](https://huggingface.co/openchat/openchat-3.5-0106).\nThese files were quantised using hardware kindly provided by [Massed Compute](https://massedcompute.com/).",
+      "description": "# Openchat 3.5 0106 - GGUF\n- Model creator: [OpenChat](https://huggingface.co/openchat)\n- Original model: [Openchat 3.5 0106](https://huggingface.co/openchat/openchat-3.5-0106)\n\n<!-- description start -->\n## Description\n\nThis repo contains GGUF format model files for [OpenChat's Openchat 3.5 0106](https://huggingface.co/openchat/openchat-3.5-0106).\n\nThese files were quantised using hardware kindly provided by [Massed Compute](https://massedcompute.com/).",
       "hw": "CPU",
       "registry": "Hugging Face",
       "license": "Apache-2.0",
@@ -149,7 +149,7 @@
     {
       "id": "hf.TheBloke.mistral-7b-openorca.Q4_K_M",
       "name": "TheBloke/Mistral-7B-OpenOrca-GGUF",
-      "description": "## Mistral 7B OpenOrca - GGUF- Model creator: [OpenOrca](https://huggingface.co/Open-Orca)\n- Original model: [Mistral 7B OpenOrca](https://huggingface.co/Open-Orca/Mistral-7B-OpenOrca)\n## Description\nThis repo contains GGUF format model files for [OpenOrca's Mistral 7B OpenOrca](https://huggingface.co/Open-Orca/Mistral-7B-OpenOrca).",
+      "description": "# Mistral 7B OpenOrca - GGUF\n- Model creator: [OpenOrca](https://huggingface.co/Open-Orca)\n- Original model: [Mistral 7B OpenOrca](https://huggingface.co/Open-Orca/Mistral-7B-OpenOrca)\n\n<!-- description start -->\n## Description\n\nThis repo contains GGUF format model files for [OpenOrca's Mistral 7B OpenOrca](https://huggingface.co/Open-Orca/Mistral-7B-OpenOrca).",
       "hw": "CPU",
       "registry": "Hugging Face",
       "license": "Apache-2.0",
@@ -159,7 +159,7 @@
     {
       "id": "hf.MaziyarPanahi.phi-2.Q4_K_M",
       "name": "MaziyarPanahi/phi-2-GGUF",
-      "description": "## [MaziyarPanahi/phi-2-GGUF](https://huggingface.co/MaziyarPanahi/phi-2-GGUF)\n- Model creator: [microsoft](https://huggingface.co/microsoft)\n- Original model: [microsoft/phi-2](https://huggingface.co/microsoft/phi-2)\n## Description\n[MaziyarPanahi/phi-2-GGUF](https://huggingface.co/MaziyarPanahi/phi-2-GGUF) contains GGUF format model files for [microsoft/phi-2](https://huggingface.co/microsoft/phi-2).",
+      "description": "# [MaziyarPanahi/phi-2-GGUF](https://huggingface.co/MaziyarPanahi/phi-2-GGUF)\n- Model creator: [microsoft](https://huggingface.co/microsoft)\n- Original model: [microsoft/phi-2](https://huggingface.co/microsoft/phi-2)\n\n## Description\n[MaziyarPanahi/phi-2-GGUF](https://huggingface.co/MaziyarPanahi/phi-2-GGUF) contains GGUF format model files for [microsoft/phi-2](https://huggingface.co/microsoft/phi-2).",
       "hw": "CPU",
       "registry": "Hugging Face",
       "license": "Apache-2.0",
@@ -169,7 +169,7 @@
     {
       "id": "hf.llmware.dragon-mistral-7b-q4_k_m",
       "name": "llmware/dragon-mistral-7b-v0",
-      "description": "## Model Card for Model ID\ndragon-mistral-7b-v0 part of the dRAGon ('Delivering RAG On ...') model series, RAG-instruct trained on top of a Mistral-7B base model.\nDRAGON models have been fine-tuned with the specific objective of fact-based question-answering over complex business and legal documents with an emphasis on reducing hallucinations and providing short, clear answers for workflow automation.",
+      "description": "# Model Card for Model ID\n\n<!-- Provide a quick summary of what the model is/does. -->\n\ndragon-mistral-7b-v0 part of the dRAGon (\"Delivering RAG On ...\") model series, RAG-instruct trained on top of a Mistral-7B base model.\n\nDRAGON models have been fine-tuned with the specific objective of fact-based question-answering over complex business and legal documents with an emphasis on reducing hallucinations and providing short, clear answers for workflow automation.",
       "hw": "CPU",
       "registry": "Hugging Face",
       "license": "Apache-2.0",
@@ -179,7 +179,7 @@
     {
       "id": "hf.MaziyarPanahi.MixTAO-7Bx2-MoE-Instruct-v7.0.Q4_K_M",
       "name": "MaziyarPanahi/MixTAO-7Bx2-MoE-Instruct-v7.0-GGUF",
-      "description": "## [MaziyarPanahi/MixTAO-7Bx2-MoE-Instruct-v7.0-GGUF](https://huggingface.co/MaziyarPanahi/MixTAO-7Bx2-MoE-Instruct-v7.0-GGUF)\n- Model creator: [zhengr](https://huggingface.co/zhengr)\n- Original model: [zhengr/MixTAO-7Bx2-MoE-Instruct-v7.0](https://huggingface.co/zhengr/MixTAO-7Bx2-MoE-Instruct-v7.0)\n## Description\n[MaziyarPanahi/MixTAO-7Bx2-MoE-Instruct-v7.0-GGUF](https://huggingface.co/MaziyarPanahi/MixTAO-7Bx2-MoE-Instruct-v7.0-GGUF) contains GGUF format model files for [zhengr/MixTAO-7Bx2-MoE-Instruct-v7.0](https://huggingface.co/zhengr/MixTAO-7Bx2-MoE-Instruct-v7.0).",
+      "description": "# [MaziyarPanahi/MixTAO-7Bx2-MoE-Instruct-v7.0-GGUF](https://huggingface.co/MaziyarPanahi/MixTAO-7Bx2-MoE-Instruct-v7.0-GGUF)\n- Model creator: [zhengr](https://huggingface.co/zhengr)\n- Original model: [zhengr/MixTAO-7Bx2-MoE-Instruct-v7.0](https://huggingface.co/zhengr/MixTAO-7Bx2-MoE-Instruct-v7.0)\n\n## Description\n[MaziyarPanahi/MixTAO-7Bx2-MoE-Instruct-v7.0-GGUF](https://huggingface.co/MaziyarPanahi/MixTAO-7Bx2-MoE-Instruct-v7.0-GGUF) contains GGUF format model files for [zhengr/MixTAO-7Bx2-MoE-Instruct-v7.0](https://huggingface.co/zhengr/MixTAO-7Bx2-MoE-Instruct-v7.0).\n",
       "hw": "CPU",
       "registry": "Hugging Face",
       "license": "Apache-2.0",