Adding model descriptions

static/models_data/Mistral-7B-Instruct-v0.1/model_detail.html

@@ -0,0 +1,6 @@
+<p>
+    This open-source model was created by <a href="https://mistral.ai/">Mistral AI<a>.
+    You can find the release blog post <a href="https://mistral.ai/news/announcing-mistral-7b/">here</a>.
+    The model is available on the huggingface hub:  <a href="https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.1">https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.1</a>.
+    The model has 7.3B parameters, and supports up to 8K token contexts.
+</p>

static/models_data/Mistral-7B-Instruct-v0.2/model_detail.html

@@ -0,0 +1,6 @@
+<p>
+    This open-source model was created by <a href="https://mistral.ai/">Mistral AI<a>.
+    You can find the release blog post <a href="https://mistral.ai/news/la-plateforme/">here</a>.
+    The model is available on the huggingface hub:  <a href="https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.2">https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.2</a>.
+    The model has 7.3B parameters, and supports up to 8K token contexts.
+</p>

static/models_data/Mistral-7B-Instruct-v0.3/model_detail.html

@@ -0,0 +1,5 @@
+<p>
+    This open-source model was created by <a href="https://mistral.ai/">Mistral AI<a>.
+    The model is available on the huggingface hub:  <a href="https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.3">https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.3</a>.
+    The model has 7.3B parameters, and supports up to 8K token contexts.
+</p>

static/models_data/Mixtral-8x22B-Instruct-v0.1/model_detail.html

@@ -0,0 +1,6 @@
+<p>
+    This open-source model was created by <a href="https://mistral.ai/">Mistral AI<a>.
+    You can find the release blog post <a href="https://mistral.ai/news/mixtral-8x22b/">here</a>.
+    The model is available on the huggingface hub:  <a href="https://huggingface.co/mistralai/Mixtral-8x22B-Instruct-v0.1">https://huggingface.co/mistralai/Mixtral-8x22B-Instruct-v0.1</a>.
+    The model has 141B total and 39B active parameters. It supports up to 64K token contexts.
+</p>

static/models_data/Mixtral-8x7B-Instruct-v0.1/model_detail.html

@@ -0,0 +1,6 @@
+<p>
+    This open-source model was created by <a href="https://mistral.ai/">Mistral AI<a>.
+    You can find the release blog post <a href="https://mistral.ai/news/mixtral-of-experts/">here</a>.
+    The model is available on the huggingface hub:  <a href="https://huggingface.co/mistralai/Mixtral-8x7B-Instruct-v0.1">https://huggingface.co/mistralai/Mixtral-8x7B-Instruct-v0.1</a>.
+    The model has 46.7B total and 12.9B active parameters. It supports up to 32K token contexts.
+</p>

static/models_data/Qwen2-72B-Instruct/model_detail.html

@@ -0,0 +1,6 @@
+<p>
+    This open-source model was created by <a href="https://qwenlm.github.io/">The Qwen Team of Alibaba cloud <a>.
+    You can find the release blog post <a href="https://qwenlm.github.io/blog/qwen2/">here</a>.
+    The model is available on the huggingface hub:  <a href="https://huggingface.co/Qwen/Qwen2-72B-Instruct">https://huggingface.co/Qwen/Qwen2-72B-Instruct</a>.
+    The 72B model was pretrained on 29 different languages, and supports up to 128K tokens.
+</p>

static/models_data/Qwen2-7B-Instruct/model_detail.html

@@ -0,0 +1,6 @@
+<p>
+    This open-source model was created by <a href="https://qwenlm.github.io/">The Qwen Team of Alibaba cloud <a>.
+    You can find the release blog post <a href="https://qwenlm.github.io/blog/qwen2/">here</a>.
+    The model is available on the huggingface hub:  <a href="https://huggingface.co/Qwen/Qwen2-7B-Instruct">https://huggingface.co/Qwen/Qwen2-7B-Instruct</a>.
+    The 7B model was pretrained on 29 different languages, and supports up to 128K tokens.
+</p>

static/models_data/command_r_plus/model_detail.html

@@ -0,0 +1,5 @@
+<p>
+    This open-source model was created by <a href="https://cohere.com/">Cohere<AI<a>.
+    The model is available on the huggingface hub:  <a href="https://huggingface.co/CohereForAI/c4ai-command-r-plus">https://huggingface.co/CohereForAI/c4ai-command-r-plus</a>.
+    The model has 104B parameters, and supports up to 128K token contexts.
+</p>

static/models_data/dummy/model_detail.html

@@ -0,0 +1,3 @@
+<p>
+    This is a dummy model. It returns a random answer (of the suggested ones) to the multiple choice question.
+</p>

static/models_data/gpt-3.5-turbo-0125/model_detail.html

@@ -0,0 +1,4 @@
+<p>
+    This proprietary model was created by <a href="https://openai.com/">OpenAI<a>.
+    You can find the release blog post <a href="https://openai.com/index/chatgpt/">here</a>.
+</p>

static/models_data/gpt-4o-0513/model_detail.html

@@ -0,0 +1,4 @@
+<p>
+    This proprietary model was created by <a href="https://openai.com/">OpenAI<a>.
+    You can find the release blog post <a href="https://openai.com/index/hello-gpt-4o/">here</a>.
+</p>

static/models_data/gpt-4o-mini-2024-07-18/model_detail.html

@@ -0,0 +1,4 @@
+<p>
+    This proprietary model was created by <a href="https://openai.com/">OpenAI<a>.
+    You can find the release blog post <a href="https://openai.com/index/gpt-4o-mini-advancing-cost-efficient-intelligence/">here</a>.
+</p>

static/models_data/llama_3.1_405b_instruct_4bit/model_detail.html

@@ -0,0 +1,8 @@
+<p>
+    This open-source model was created by <a href="https://ai.meta.com/">Meta AI</a>.
+    You can find the release blog post <a href="https://ai.meta.com/blog/meta-llama-3-1/">here</a>.
+    The 16bit precision model is available on the huggingface hub:  <a href="https://huggingface.co/meta-llama/Meta-Llama-3.1-405B-Instruct">https://huggingface.co/meta-llama/Meta-Llama-3.1-405B-Instruct</a>.
+    Due to computational constrains we use the 4bit quantized version, which is also available on the huggingfacehub: <a href="unsloth/Meta-Llama-3.1-405B-Instruct-bnb-4bit">unsloth/Meta-Llama-3.1-405B-Instruct-bnb-4bit</a>.
+    It is relevant to note that we compared with a 16bit version hosted by <a href="https://www.together.ai/">TogetherAI</a> on a subset of problems that fall in the 4k tokens limit defined by the TogetherAI API, and we did not see drastic changes in performance.
+    The 405B model was pretrained on 15 trillion tokens spanning 8 different languages, and supports up to 128K token contexts.
+</p>

static/models_data/llama_3.1_70b_instruct/model_detail.html

@@ -0,0 +1,6 @@
+<p>
+    This open-source model was created by <a href="https://ai.meta.com/">Meta AI</a>.
+    You can find the release blog post <a href="https://ai.meta.com/blog/meta-llama-3-1/">here</a>.
+    The model is available on the huggingface hub:  <a href="https://huggingface.co/meta-llama/Meta-Llama-3.1-70B-Instruct">https://huggingface.co/meta-llama/Meta-Llama-3.1-70B-Instruct</a>.
+    The 70B model was pretrained on 15 trillion tokens spanning 8 different languages, and supports up to 128K token contexts.
+</p>

static/models_data/llama_3.1_8b_instruct/model_detail.html

@@ -0,0 +1,6 @@
+<p>
+    This open-source model was created by <a href="https://ai.meta.com/">Meta AI</a>.
+    You can find the release blog post <a href="https://ai.meta.com/blog/meta-llama-3-1/">here</a>.
+    The model is available on the huggingface hub:  <a href="https://huggingface.co/meta-llama/Meta-Llama-3.1-8B-Instruct">https://huggingface.co/meta-llama/Meta-Llama-3.1-8B-Instruct</a>.
+    The 70B model was pretrained on 15 trillion tokens spanning 8 different languages, and supports up to 128K token contexts.
+</p>

static/models_data/llama_3_70b_instruct/model_detail.html

@@ -0,0 +1,6 @@
+<p>
+    This open-source model was created by <a href="https://ai.meta.com/">Meta AI</a>.
+    You can find the release blog post <a href="https://ai.meta.com/blog/meta-llama-3/">here</a>.
+    The model is available on the huggingface hub:  <a href="https://huggingface.co/meta-llama/Meta-Llama-3-70B-Instruct">https://huggingface.co/meta-llama/Meta-Llama-3-70B-Instruct</a>.
+    The 70B model was pretrained on 15 trillion tokens spanning 30 different languages in sequences of 8,192 tokens.
+</p>

static/models_data/llama_3_8b_instruct/model_detail.html

@@ -0,0 +1,6 @@
+<p>
+    This open-source model was created by <a href="https://ai.meta.com/">Meta AI</a>.
+    You can find the release blog post <a href="https://ai.meta.com/blog/meta-llama-3/">here</a>.
+    The model is available on the huggingface hub:  <a href="https://huggingface.co/meta-llama/Meta-Llama-3-8B-Instruct">https://huggingface.co/meta-llama/Meta-Llama-3-8B-Instruct</a>.
+    The 8B model was pretrained on 15 trillion tokens spanning 30 different languages in sequences of 8,192 tokens.
+</p>

static/models_data/phi-3-medium-128k-instruct/model_detail.html

@@ -0,0 +1,6 @@
+<p>
+    This open-source model was created by <a href="https://www.microsoft.com/">Microsoft<a>.
+    You can find the release blog post <a href="https://azure.microsoft.com/en-us/blog/introducing-phi-3-redefining-whats-possible-with-slms/">here</a>.
+    The model is available on the huggingface hub:  <a href="https://huggingface.co/microsoft/Phi-3-medium-128k-instruct">https://huggingface.co/microsoft/Phi-3-medium-128k-instruct</a>.
+    The model has 14B parameters, and supports up to 128K token contexts.
+</p>

static/models_data/phi-3-mini-128k-instruct/model_detail.html

@@ -0,0 +1,6 @@
+<p>
+    This open-source model was created by <a href="https://www.microsoft.com/">Microsoft<a>.
+    You can find the release blog post <a href="https://azure.microsoft.com/en-us/blog/introducing-phi-3-redefining-whats-possible-with-slms/">here</a>.
+    The model is available on the huggingface hub:  <a href="https://huggingface.co/microsoft/Phi-3-mini-128k-instruct">https://huggingface.co/microsoft/Phi-3-mini-128k-instruct</a>.
+    The model has 3.8B parameters, and supports up to 128K token contexts.
+</p>