PyPI - crfm-helm - Versions diffs - 0.5.2__py3-none-any.whl → 0.5.3__py3-none-any.whl - Mend - Supply Chain Defender

crfm-helm 0.5.2py3-none-any.whl → 0.5.3py3-none-any.whl

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.

Potentially problematic release.

This version of crfm-helm might be problematic. Click here for more details.

Files changed (184) hide show

helm/config/model_metadata.yaml CHANGED Viewed

@@ -31,50 +31,41 @@ models:
   # AI21 Labs
-  - name: ai21/j1-jumbo # DEPRECATED
+  - name: ai21/j1-jumbo
     display_name: J1-Jumbo v1 (178B)
     description: Jurassic-1 Jumbo (178B parameters) ([docs](https://studio.ai21.com/docs/jurassic1-language-models/), [tech report](https://uploads-ssl.webflow.com/60fd4503684b466578c0d307/61138924626a6981ee09caf6_jurassic_tech_paper.pdf)).
     creator_organization_name: AI21 Labs
     access: limited
     num_parameters: 178000000000
     release_date: 2021-08-11
-    tags: [TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG]
+    tags: [DEPRECATED_MODEL_TAG, TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG]
-  - name: ai21/j1-large # DEPRECATED
+  - name: ai21/j1-large
     display_name: J1-Large v1 (7.5B)
     description: Jurassic-1 Large (7.5B parameters) ([docs](https://studio.ai21.com/docs/jurassic1-language-models/), [tech report](https://uploads-ssl.webflow.com/60fd4503684b466578c0d307/61138924626a6981ee09caf6_jurassic_tech_paper.pdf)).
     creator_organization_name: AI21 Labs
     access: limited
     num_parameters: 7500000000
     release_date: 2021-08-11
-    tags: [TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG]
+    tags: [DEPRECATED_MODEL_TAG, TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG]
-  - name: ai21/j1-grande # DEPRECATED
+  - name: ai21/j1-grande
     display_name: J1-Grande v1 (17B)
     description: Jurassic-1 Grande (17B parameters) with a "few tweaks" to the training process ([docs](https://studio.ai21.com/docs/jurassic1-language-models/), [tech report](https://uploads-ssl.webflow.com/60fd4503684b466578c0d307/61138924626a6981ee09caf6_jurassic_tech_paper.pdf)).
     creator_organization_name: AI21 Labs
     access: limited
     num_parameters: 17000000000
     release_date: 2022-05-03
-    tags: [TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG]
+    tags: [DEPRECATED_MODEL_TAG, TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG]
-  - name: ai21/j1-grande-v2-beta # DEPRECATED
+  - name: ai21/j1-grande-v2-beta
     display_name: J1-Grande v2 beta (17B)
     description: Jurassic-1 Grande v2 beta (17B parameters)
     creator_organization_name: AI21 Labs
     access: limited
     num_parameters: 17000000000
     release_date: 2022-10-28
-    tags: [TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG]
-  - name: ai21/j2-jumbo
-    display_name: Jurassic-2 Jumbo (178B)
-    description: Jurassic-2 Jumbo (178B parameters) ([docs](https://www.ai21.com/blog/introducing-j2))
-    creator_organization_name: AI21 Labs
-    access: limited
-    num_parameters: 178000000000
-    release_date: 2023-03-09
-    tags: [TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG]
+    tags: [DEPRECATED_MODEL_TAG, TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG]
   - name: ai21/j2-large
     display_name: Jurassic-2 Large (7.5B)
@@ -83,7 +74,7 @@ models:
     access: limited
     num_parameters: 7500000000
     release_date: 2023-03-09
-    tags: [TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG]
+    tags: [TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
   - name: ai21/j2-grande
     display_name: Jurassic-2 Grande (17B)
@@ -92,13 +83,48 @@ models:
     access: limited
     num_parameters: 17000000000
     release_date: 2023-03-09
-    tags: [TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG]
+    tags: [TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+  - name: ai21/j2-jumbo
+    display_name: Jurassic-2 Jumbo (178B)
+    description: Jurassic-2 Jumbo (178B parameters) ([docs](https://www.ai21.com/blog/introducing-j2))
+    creator_organization_name: AI21 Labs
+    access: limited
+    num_parameters: 178000000000
+    release_date: 2023-03-09
+    tags: [TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
   # TODO(1524): Change AI21 model names
   # - j2-jumbo -> j2-ultra
   # - j2-grande -> j2-mid
   # - j2-large -> j2-light
+  - name: ai21/jamba-instruct
+    display_name: Jamba Instruct
+    description: Jamba Instruct is an instruction tuned version of Jamba, which uses a hybrid Transformer-Mamba mixture-of-experts (MoE) architecture that interleaves blocks of Transformer and Mamba layers. ([blog](https://www.ai21.com/blog/announcing-jamba-instruct))
+    creator_organization_name: AI21 Labs
+    access: limited
+    num_parameters: 52000000000
+    release_date: 2024-05-02
+    tags: [TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+  - name: ai21/jamba-1.5-mini
+    display_name: Jamba 1.5 Mini
+    description: Jamba 1.5 Mini is a long-context, hybrid SSM-Transformer instruction following foundation model that is optimized for function calling, structured output, and grounded generation. ([blog](https://www.ai21.com/blog/announcing-jamba-model-family))
+    creator_organization_name: AI21 Labs
+    access: open
+    num_parameters: 51600000000
+    release_date: 2024-08-22
+    tags: [TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+  - name: ai21/jamba-1.5-large
+    display_name: Jamba 1.5 Large
+    description: Jamba 1.5 Large is a long-context, hybrid SSM-Transformer instruction following foundation model that is optimized for function calling, structured output, and grounded generation. ([blog](https://www.ai21.com/blog/announcing-jamba-model-family))
+    creator_organization_name: AI21 Labs
+    access: open
+    num_parameters: 399000000000
+    release_date: 2024-08-22
+    tags: [TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
   # AI Singapore
   - name: aisingapore/sea-lion-7b
@@ -117,7 +143,7 @@ models:
     access: open
     num_parameters: 7000000000
     release_date: 2023-02-24
-    tags: [TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG]
+    tags: [TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
   # Aleph Alpha
@@ -272,7 +298,14 @@ models:
     release_date: 2024-03-04  # https://www.anthropic.com/news/claude-3-family
     tags: [ANTHROPIC_CLAUDE_3_MODEL_TAG, TEXT_MODEL_TAG, VISION_LANGUAGE_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
-  # DEPRECATED: Please do not use.
+  - name: anthropic/claude-3-5-sonnet-20240620
+    display_name: Claude 3.5 Sonnet (20240620)
+    description: Claude 3.5 Sonnet is a Claude 3 family model which outperforms Claude 3 Opus while operating faster and at a lower cost. ([blog](https://www.anthropic.com/news/claude-3-5-sonnet))
+    creator_organization_name: Anthropic
+    access: limited
+    release_date: 2024-06-20
+    tags: [ANTHROPIC_CLAUDE_3_MODEL_TAG, TEXT_MODEL_TAG, VISION_LANGUAGE_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
   - name: anthropic/stanford-online-all-v4-s3
     display_name: Anthropic-LM v4-s3 (52B)
     description: A 52B parameter language model, trained using reinforcement learning from human feedback [paper](https://arxiv.org/pdf/2204.05862.pdf).
@@ -280,7 +313,7 @@ models:
     access: closed
     num_parameters: 52000000000
     release_date: 2021-12-01
-    tags: [TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG, ABLATION_MODEL_TAG]
+    tags: [DEPRECATED_MODEL_TAG, TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG, ABLATION_MODEL_TAG]
@@ -401,16 +434,16 @@ models:
     access: limited
     num_parameters: 52400000000
     release_date: 2022-06-09
-    tags: [TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG]
+    tags: [DEPRECATED_MODEL_TAG, TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG]
-  - name: cohere/large-20220720 # DEPRECATED
+  - name: cohere/large-20220720
     display_name: Cohere large v20220720 (13.1B)
     description: Cohere large v20220720 (13.1B parameters), which is deprecated by Cohere as of December 2, 2022.
     creator_organization_name: Cohere
     access: limited
     num_parameters: 13100000000
     release_date: 2022-07-20
-    tags: [TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG]
+    tags: [DEPRECATED_MODEL_TAG, TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG]
   - name: cohere/medium-20220720
     display_name: Cohere medium v20220720 (6.1B)
@@ -419,16 +452,16 @@ models:
     access: limited
     num_parameters: 6100000000
     release_date: 2022-07-20
-    tags: [TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG]
+    tags: [DEPRECATED_MODEL_TAG, TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG]
-  - name: cohere/small-20220720 # DEPRECATED
+  - name: cohere/small-20220720
     display_name: Cohere small v20220720 (410M)
     description: Cohere small v20220720 (410M parameters), which is deprecated by Cohere as of December 2, 2022.
     creator_organization_name: Cohere
     access: limited
     num_parameters: 410000000
     release_date: 2022-07-20
-    tags: [TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG]
+    tags: [DEPRECATED_MODEL_TAG, TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG]
   - name: cohere/xlarge-20221108
     display_name: Cohere xlarge v20221108 (52.4B)
@@ -437,34 +470,34 @@ models:
     access: limited
     num_parameters: 52400000000
     release_date: 2022-11-08
-    tags: [TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG]
+    tags: [DEPRECATED_MODEL_TAG, TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG]
-  - name: cohere/medium-20221108 # DEPRECATED
+  - name: cohere/medium-20221108
     display_name: Cohere medium v20221108 (6.1B)
     description: Cohere medium v20221108 (6.1B parameters)
     creator_organization_name: Cohere
     access: limited
     num_parameters: 6100000000
     release_date: 2022-11-08
-    tags: [TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG]
+    tags: [DEPRECATED_MODEL_TAG, TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG]
-  - name: cohere/command-medium-beta # DEPRECATED
+  - name: cohere/command-medium-beta
     display_name: Command beta (6.1B)
-    description: Cohere Command beta (6.1B parameters) is fine-tuned from the medium model to respond well with instruction-like prompts ([details](https://docs.cohere.ai/docs/command-beta)).
+    description: Command beta (6.1B parameters) is fine-tuned from the medium model to respond well with instruction-like prompts ([details](https://docs.cohere.ai/docs/command-beta)).
     creator_organization_name: Cohere
     access: limited
     num_parameters: 6100000000
     release_date: 2022-11-08
-    tags: [TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+    tags: [DEPRECATED_MODEL_TAG, TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
-  - name: cohere/command-xlarge-beta # DEPRECATED
+  - name: cohere/command-xlarge-beta
     display_name: Command beta (52.4B)
-    description: Cohere Command beta (52.4B parameters) is fine-tuned from the XL model to respond well with instruction-like prompts ([details](https://docs.cohere.ai/docs/command-beta)).
+    description: Command beta (52.4B parameters) is fine-tuned from the XL model to respond well with instruction-like prompts ([details](https://docs.cohere.ai/docs/command-beta)).
     creator_organization_name: Cohere
     access: limited
     num_parameters: 52400000000
     release_date: 2022-11-08
-    tags: [TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+    tags: [DEPRECATED_MODEL_TAG, TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
   - name: cohere/command
     display_name: Command
@@ -837,8 +870,7 @@ models:
   - name: google/gemma-2b
     display_name: Gemma (2B)
-    # TODO: Fill in Gemma description.
-    description: TBD
+    description: Gemma is a family of lightweight, open models built from the research and technology that Google used to create the Gemini models. ([model card](https://www.kaggle.com/models/google/gemma), [blog post](https://blog.google/technology/developers/gemma-open-models/))
     creator_organization_name: Google
     access: open
     release_date: 2024-02-21
@@ -846,8 +878,7 @@ models:
   - name: google/gemma-2b-it
     display_name: Gemma Instruct (2B)
-    # TODO: Fill in Gemma description.
-    description: TBD
+    description: Gemma is a family of lightweight, open models built from the research and technology that Google used to create the Gemini models. ([model card](https://www.kaggle.com/models/google/gemma), [blog post](https://blog.google/technology/developers/gemma-open-models/))
     creator_organization_name: Google
     access: open
     release_date: 2024-02-21
@@ -855,8 +886,7 @@ models:
   - name: google/gemma-7b
     display_name: Gemma (7B)
-    # TODO: Fill in Gemma description.
-    description: TBD
+    description: Gemma is a family of lightweight, open models built from the research and technology that Google used to create the Gemini models. ([model card](https://www.kaggle.com/models/google/gemma), [blog post](https://blog.google/technology/developers/gemma-open-models/))
     creator_organization_name: Google
     access: open
     release_date: 2024-02-21
@@ -864,12 +894,42 @@ models:
   - name: google/gemma-7b-it
     display_name: Gemma Instruct (7B)
-    # TODO: Fill in Gemma description.
-    description: TBD
+    description: Gemma is a family of lightweight, open models built from the research and technology that Google used to create the Gemini models. ([model card](https://www.kaggle.com/models/google/gemma), [blog post](https://blog.google/technology/developers/gemma-open-models/))
     creator_organization_name: Google
     access: open
     release_date: 2024-02-21
-    # TODO: Add OUTPUT_FORMAT_INSTRUCTIONS_TAG tag
+    tags: [TEXT_MODEL_TAG, GOOGLE_GEMMA_INSTRUCT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+  - name: google/gemma-2-9b
+    display_name: Gemma 2 (9B)
+    description: Gemma is a family of lightweight, open models built from the research and technology that Google used to create the Gemini models. ([model card](https://www.kaggle.com/models/google/gemma), [blog post](https://blog.google/technology/developers/google-gemma-2/))
+    creator_organization_name: Google
+    access: open
+    release_date: 2024-06-27
+    tags: [TEXT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG]
+  - name: google/gemma-2-9b-it
+    display_name: Gemma 2 Instruct (9B)
+    description: Gemma is a family of lightweight, open models built from the research and technology that Google used to create the Gemini models. ([model card](https://www.kaggle.com/models/google/gemma), [blog post](https://blog.google/technology/developers/google-gemma-2/))
+    creator_organization_name: Google
+    access: open
+    release_date: 2024-06-27
+    tags: [TEXT_MODEL_TAG, GOOGLE_GEMMA_INSTRUCT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+  - name: google/gemma-2-27b
+    display_name: Gemma 2 (27B)
+    description: Gemma is a family of lightweight, open models built from the research and technology that Google used to create the Gemini models. ([model card](https://www.kaggle.com/models/google/gemma), [blog post](https://blog.google/technology/developers/google-gemma-2/))
+    creator_organization_name: Google
+    access: open
+    release_date: 2024-06-27
+    tags: [TEXT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG]
+  - name: google/gemma-2-27b-it
+    display_name: Gemma 2 Instruct (27B)
+    description: Gemma is a family of lightweight, open models built from the research and technology that Google used to create the Gemini models. ([model card](https://www.kaggle.com/models/google/gemma), [blog post](https://blog.google/technology/developers/google-gemma-2/))
+    creator_organization_name: Google
+    access: open
+    release_date: 2024-06-27
     tags: [TEXT_MODEL_TAG, GOOGLE_GEMMA_INSTRUCT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
   - name: google/paligemma-3b-mix-224
@@ -1361,7 +1421,7 @@ models:
   - name: meta/llama-3-8b
     display_name: Llama 3 (8B)
-    description: Llama 3 is a family of language models that have been trained on more than 15 trillion tokens, and use Grouped-Query Attention (GQA) for improved inference scalability.
+    description: Llama 3 is a family of language models that have been trained on more than 15 trillion tokens, and use Grouped-Query Attention (GQA) for improved inference scalability. ([paper](https://ai.meta.com/research/publications/the-llama-3-herd-of-models/)
     creator_organization_name: Meta
     access: open
     num_parameters: 8000000000
@@ -1370,16 +1430,43 @@ models:
   - name: meta/llama-3-70b
     display_name: Llama 3 (70B)
-    description: Llama 3 is a family of language models that have been trained on more than 15 trillion tokens, and use Grouped-Query Attention (GQA) for improved inference scalability.
+    description: Llama 3 is a family of language models that have been trained on more than 15 trillion tokens, and use Grouped-Query Attention (GQA) for improved inference scalability. ([paper](https://ai.meta.com/research/publications/the-llama-3-herd-of-models/)
     creator_organization_name: Meta
     access: open
     num_parameters: 70000000000
     release_date: 2024-04-18
     tags: [TEXT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG]
+  - name: meta/llama-3.1-8b-instruct-turbo
+    display_name: Llama 3.1 Instruct Turbo (8B)
+    description: Llama 3.1 (8B) is part of the Llama 3 family of dense Transformer models that that natively support multilinguality, coding, reasoning, and tool usage. ([paper](https://ai.meta.com/research/publications/the-llama-3-herd-of-models/), [blog](https://ai.meta.com/blog/meta-llama-3-1/)) Turbo is Together's implementation, providing a near negligible difference in quality from the reference implementation with faster performance and lower cost, currently using FP8 quantization. ([blog](https://www.together.ai/blog/llama-31-quality))
+    creator_organization_name: Meta
+    access: open
+    num_parameters: 8000000000
+    release_date: 2024-07-23
+    tags: [TEXT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG]
+  - name: meta/llama-3.1-70b-instruct-turbo
+    display_name: Llama 3.1 Instruct Turbo (70B)
+    description: Llama 3.1 (70B) is part of the Llama 3 family of dense Transformer models that that natively support multilinguality, coding, reasoning, and tool usage. ([paper](https://ai.meta.com/research/publications/the-llama-3-herd-of-models/), [blog](https://ai.meta.com/blog/meta-llama-3-1/)) Turbo is Together's implementation, providing a near negligible difference in quality from the reference implementation with faster performance and lower cost, currently using FP8 quantization. ([blog](https://www.together.ai/blog/llama-31-quality))
+    creator_organization_name: Meta
+    access: open
+    num_parameters: 70000000000
+    release_date: 2024-07-23
+    tags: [TEXT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG]
+  - name: meta/llama-3.1-405b-instruct-turbo
+    display_name: Llama 3.1 Instruct Turbo (405B)
+    description: Llama 3.1 (405B) is part of the Llama 3 family of dense Transformer models that that natively support multilinguality, coding, reasoning, and tool usage. ([paper](https://ai.meta.com/research/publications/the-llama-3-herd-of-models/), [blog](https://ai.meta.com/blog/meta-llama-3-1/)) Turbo is Together's implementation, providing a near negligible difference in quality from the reference implementation with faster performance and lower cost, currently using FP8 quantization. ([blog](https://www.together.ai/blog/llama-31-quality))
+    creator_organization_name: Meta
+    access: open
+    num_parameters: 405000000000
+    release_date: 2024-07-23
+    tags: [TEXT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG]
   - name: meta/llama-3-8b-chat
     display_name: Llama 3 Instruct (8B)
-    description: Llama 3 is a family of language models that have been trained on more than 15 trillion tokens, and use Grouped-Query Attention (GQA) for improved inference scalability. It used SFT, rejection sampling, PPO and DPO for post-training.
+    description: Llama 3 is a family of language models that have been trained on more than 15 trillion tokens, and use Grouped-Query Attention (GQA) for improved inference scalability. It used SFT, rejection sampling, PPO and DPO for post-training. ([paper](https://ai.meta.com/research/publications/the-llama-3-herd-of-models/)
     creator_organization_name: Meta
     access: open
     num_parameters: 8000000000
@@ -1388,7 +1475,7 @@ models:
   - name: meta/llama-3-70b-chat
     display_name: Llama 3 Instruct (70B)
-    description: Llama 3 is a family of language models that have been trained on more than 15 trillion tokens, and use Grouped-Query Attention (GQA) for improved inference scalability. It used SFT, rejection sampling, PPO and DPO for post-training.
+    description: Llama 3 is a family of language models that have been trained on more than 15 trillion tokens, and use Grouped-Query Attention (GQA) for improved inference scalability. It used SFT, rejection sampling, PPO and DPO for post-training. ([paper](https://ai.meta.com/research/publications/the-llama-3-herd-of-models/)
     creator_organization_name: Meta
     access: open
     num_parameters: 70000000000
@@ -1413,6 +1500,16 @@ models:
     release_date: 2024-04-18
     tags: [TEXT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+  - name: meta/llama-guard-3-8b
+    display_name: Llama Guard 3 (8B)
+    description: Llama Guard 3 is an 8B parameter Llama 3.1-based LLM safeguard model. Similar to Llama Guard, it can be used for classifying content in both LLM inputs (prompt classification) and in LLM responses (response classification). It acts as an LLM – it generates text in its output that indicates whether a given prompt or response is safe or unsafe, and if unsafe, it also lists the content categories violated.
+    creator_organization_name: Meta
+    access: open
+    num_parameters: 8000000000
+    release_date: 2024-07-23
+    tags: [TEXT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
@@ -1424,7 +1521,7 @@ models:
     access: closed
     num_parameters: 530000000000
     release_date: 2022-01-28
-    tags: []  # deprecated text model
+    tags: [DEPRECATED_MODEL_TAG, TEXT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG]
   - name: microsoft/TNLGv2_7B
     display_name: TNLG v2 (6.7B)
@@ -1433,7 +1530,7 @@ models:
     access: closed
     num_parameters: 6700000000
     release_date: 2022-01-28
-    tags: []  # deprecated text model
+    tags: [DEPRECATED_MODEL_TAG, TEXT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG]
   - name: microsoft/llava-1.5-7b-hf
     display_name: LLaVA 1.5 (7B)
@@ -1507,6 +1604,24 @@ models:
     release_date: 2023-10-05
     tags: [TEXT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG]
+  - name: microsoft/phi-3-small-8k-instruct
+    display_name: Phi-3 (7B)
+    description: Phi-3-Small-8K-Instruct is a lightweight model trained with synthetic data and filtered publicly available website data with a focus on high-quality and reasoning dense properties. ([paper](https://arxiv.org/abs/2404.14219), [blog](https://azure.microsoft.com/en-us/blog/new-models-added-to-the-phi-3-family-available-on-microsoft-azure/))
+    creator_organization_name: Microsoft
+    access: open
+    num_parameters: 7000000000
+    release_date: 2024-05-21
+    tags: [TEXT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+  - name: microsoft/phi-3-medium-4k-instruct
+    display_name: Phi-3 (14B)
+    description: Phi-3-Medium-4K-Instruct is a lightweight model trained with synthetic data and filtered publicly available website data with a focus on high-quality and reasoning dense properties. ([paper](https://arxiv.org/abs/2404.14219), [blog](https://azure.microsoft.com/en-us/blog/new-models-added-to-the-phi-3-family-available-on-microsoft-azure/))
+    creator_organization_name: Microsoft
+    access: open
+    num_parameters: 14000000000
+    release_date: 2024-05-21
+    tags: [TEXT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
   # KAIST AI
   - name: kaistai/prometheus-vision-13b-v1.0-hf
     display_name: LLaVA + Vicuna-v1.5 (13B)
@@ -1663,7 +1778,7 @@ models:
     num_parameters: 46700000000
     # Blog post: https://mistral.ai/news/mixtral-of-experts/
     release_date: 2023-12-11
-    tags: [TEXT_MODEL_TAG, PARTIAL_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG, MISTRAL_MODEL_TAG]
+    tags: [TEXT_MODEL_TAG, PARTIAL_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
   - name: mistralai/mixtral-8x22b
     display_name: Mixtral (8x22B)
@@ -1694,13 +1809,11 @@ models:
   - name: mistralai/mistral-small-2402
     display_name: Mistral Small (2402)
-    # TODO: Fill in description
-    description: TBD
+    description: Mistral Small is a multilingual model with a 32K tokens context window and function-calling capabilities. ([blog](https://mistral.ai/news/mistral-large/))
     creator_organization_name: Mistral AI
     access: limited
-    # Blog post: https://mistral.ai/news/mistral-large/
     release_date: 2023-02-26
-    tags: [TEXT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG, MISTRAL_MODEL_TAG]
+    tags: [TEXT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
   - name: mistralai/mistral-medium-2312
     display_name: Mistral Medium (2312)
@@ -1708,18 +1821,32 @@ models:
     creator_organization_name: Mistral AI
     access: limited
     release_date: 2023-12-11
-    tags: [TEXT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG, MISTRAL_MODEL_TAG]
+    tags: [TEXT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
   - name: mistralai/mistral-large-2402
     display_name: Mistral Large (2402)
-    # TODO: Fill in description
-    description: TBD
+    description: Mistral Large is a multilingual model with a 32K tokens context window and function-calling capabilities. ([blog](https://mistral.ai/news/mistral-large/))
     creator_organization_name: Mistral AI
     access: limited
-    # Blog post: https://mistral.ai/news/mistral-large/
     release_date: 2023-02-26
-    tags: [TEXT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG, MISTRAL_MODEL_TAG]
+    tags: [TEXT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+  - name: mistralai/mistral-large-2407
+    display_name: Mistral Large 2 (2407)
+    description: Mistral Large 2 is a 123 billion parameter model that has a 128k context window and supports dozens of languages and 80+ coding languages. ([blog](https://mistral.ai/news/mistral-large-2407/))
+    creator_organization_name: Mistral AI
+    access: open
+    num_parameters: 123000000000
+    release_date: 2023-07-24
+    tags: [TEXT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+  - name: mistralai/open-mistral-nemo-2407
+    display_name: Mistral NeMo (2402)
+    description: Mistral NeMo is a multilingual 12B model with a large context window of 128K tokens. ([blog](https://mistral.ai/news/mistral-nemo/))
+    creator_organization_name: Mistral AI
+    access: open
+    release_date: 2024-07-18
+    tags: [TEXT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
   # MosaicML
   - name: mosaicml/mpt-7b
@@ -1798,7 +1925,13 @@ models:
     release_date: 2019-09-17 # paper date
     tags: [TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG, BUGGY_TEMP_0_TAG]
+  - name: nvidia/nemotron-4-340b-instruct
+    display_name: Nemotron-4 Instruct (340B)
+    description: Nemotron-4 Instruct (340B) is an open weights model sized to fit on a single DGX H100 with 8 GPUs when deployed in FP8 precision. 98% of the data used for model alignment was synthetically generated ([paper](https://arxiv.org/abs/2406.11704)).
+    creator_organization_name: NVIDIA
+    access: open
+    release_date: 2024-06-17
+    tags: [TEXT_MODEL_TAG, PARTIAL_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
   # OpenAI
@@ -1836,97 +1969,95 @@ models:
   # DEPRECATED: Announced on July 06 2023 that these models will be shut down on January 04 2024.
-  - name: openai/davinci # DEPRECATED
+  - name: openai/davinci
     display_name: davinci (175B)
     description: Original GPT-3 (175B parameters) autoregressive language model ([paper](https://arxiv.org/pdf/2005.14165.pdf), [docs](https://beta.openai.com/docs/model-index-for-researchers)).
     creator_organization_name: OpenAI
     access: limited
     num_parameters: 175000000000
     release_date: 2020-05-28
-    tags: [TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG]
+    tags: [DEPRECATED_MODEL_TAG, TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG]
-  - name: openai/curie # DEPRECATED
+  - name: openai/curie
     display_name: curie (6.7B)
     description: Original GPT-3 (6.7B parameters) autoregressive language model ([paper](https://arxiv.org/pdf/2005.14165.pdf), [docs](https://beta.openai.com/docs/model-index-for-researchers)).
     creator_organization_name: OpenAI
     access: limited
     num_parameters: 6700000000
     release_date: 2020-05-28
-    tags: [TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG]
+    tags: [DEPRECATED_MODEL_TAG, TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG]
-  - name: openai/babbage # DEPRECATED
+  - name: openai/babbage
     display_name: babbage (1.3B)
     description: Original GPT-3 (1.3B parameters) autoregressive language model ([paper](https://arxiv.org/pdf/2005.14165.pdf), [docs](https://beta.openai.com/docs/model-index-for-researchers)).
     creator_organization_name: OpenAI
     access: limited
     num_parameters: 1300000000
     release_date: 2020-05-28
-    tags: [TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG]
+    tags: [DEPRECATED_MODEL_TAG, TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG]
-  - name: openai/ada # DEPRECATED
+  - name: openai/ada
     display_name: ada (350M)
     description: Original GPT-3 (350M parameters) autoregressive language model ([paper](https://arxiv.org/pdf/2005.14165.pdf), [docs](https://beta.openai.com/docs/model-index-for-researchers)).
     creator_organization_name: OpenAI
     access: limited
     num_parameters: 350000000
     release_date: 2020-05-28
-    tags: [TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG]
+    tags: [DEPRECATED_MODEL_TAG, TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG]
-  - name: openai/text-davinci-003 # DEPRECATED
+  - name: openai/text-davinci-003
     display_name: GPT-3.5 (text-davinci-003)
     description: text-davinci-003 model that involves reinforcement learning (PPO) with reward models. Derived from text-davinci-002 ([docs](https://beta.openai.com/docs/model-index-for-researchers)).
     creator_organization_name: OpenAI
     access: limited
     num_parameters: 175000000000
     release_date: 2022-11-28
-    tags: [TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+    tags: [DEPRECATED_MODEL_TAG, TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
-  # TODO: text-davinci-002 supports insertion. Support insertion in our framework.
-  #       https://github.com/stanford-crfm/benchmarking/issues/359
-  - name: openai/text-davinci-002 # DEPRECATED
+  - name: openai/text-davinci-002
     display_name: GPT-3.5 (text-davinci-002)
     description: text-davinci-002 model that involves supervised fine-tuning on human-written demonstrations. Derived from code-davinci-002 ([docs](https://beta.openai.com/docs/model-index-for-researchers)).
     creator_organization_name: OpenAI
     access: limited
     num_parameters: 175000000000
     release_date: 2022-01-27
-    tags: [TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG]
+    tags: [DEPRECATED_MODEL_TAG, TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG]
-  - name: openai/text-davinci-001 # DEPRECATED
+  - name: openai/text-davinci-001
     display_name: GPT-3.5 (text-davinci-001)
     description: text-davinci-001 model that involves supervised fine-tuning on human-written demonstrations ([docs](https://beta.openai.com/docs/model-index-for-researchers)).
     creator_organization_name: OpenAI
     access: limited
     num_parameters: 175000000000
     release_date: 2022-01-27
-    tags: [TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG]
+    tags: [DEPRECATED_MODEL_TAG, TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG]
-  - name: openai/text-curie-001 # DEPRECATED
+  - name: openai/text-curie-001
     display_name: text-curie-001
     description: text-curie-001 model that involves supervised fine-tuning on human-written demonstrations ([docs](https://beta.openai.com/docs/model-index-for-researchers)).
     creator_organization_name: OpenAI
     access: limited
     num_parameters: 6700000000
     release_date: 2022-01-27
-    tags: [TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG]
+    tags: [DEPRECATED_MODEL_TAG, TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG]
-  - name: openai/text-babbage-001 # DEPRECATED
+  - name: openai/text-babbage-001
     display_name: text-babbage-001
     description: text-babbage-001 model that involves supervised fine-tuning on human-written demonstrations ([docs](https://beta.openai.com/docs/model-index-for-researchers)).
     creator_organization_name: OpenAI
     access: limited
     num_parameters: 1300000000
     release_date: 2022-01-27
-    tags: [TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG]
+    tags: [DEPRECATED_MODEL_TAG, TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG]
-  - name: openai/text-ada-001 # DEPRECATED
+  - name: openai/text-ada-001
     display_name: text-ada-001
     description: text-ada-001 model that involves supervised fine-tuning on human-written demonstrations ([docs](https://beta.openai.com/docs/model-index-for-researchers)).
     creator_organization_name: OpenAI
     access: limited
     num_parameters: 350000000
     release_date: 2022-01-27
-    tags: [TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG]
+    tags: [DEPRECATED_MODEL_TAG, TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG]
   ## GPT 3.5 Turbo Models
@@ -2044,16 +2175,32 @@ models:
     creator_organization_name: OpenAI
     access: limited
     release_date: 2024-04-09
-    tags: [TEXT_MODEL_TAG, OPENAI_CHATGPT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+    tags: [TEXT_MODEL_TAG, VISION_LANGUAGE_MODEL_TAG, OPENAI_CHATGPT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
   - name: openai/gpt-4o-2024-05-13
     display_name: GPT-4o (2024-05-13)
-    description: GPT-4o (2024-05-13) is a large multimodal model that accepts as input any combination of text, audio, and image and generates any combination of text, audio, and image outputs.
+    description: GPT-4o (2024-05-13) is a large multimodal model that accepts as input any combination of text, audio, and image and generates any combination of text, audio, and image outputs. ([blog](https://openai.com/index/hello-gpt-4o/))
     creator_organization_name: OpenAI
     access: limited
     release_date: 2024-04-09
     tags: [TEXT_MODEL_TAG, VISION_LANGUAGE_MODEL_TAG, OPENAI_CHATGPT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+  - name: openai/gpt-4o-2024-08-06
+    display_name: GPT-4o (2024-08-06)
+    description: GPT-4o (2024-08-06) is a large multimodal model that accepts as input any combination of text, audio, and image and generates any combination of text, audio, and image outputs. ([blog](https://openai.com/index/introducing-structured-outputs-in-the-api/))
+    creator_organization_name: OpenAI
+    access: limited
+    release_date: 2024-08-06
+    tags: [TEXT_MODEL_TAG, VISION_LANGUAGE_MODEL_TAG, OPENAI_CHATGPT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+  - name: openai/gpt-4o-mini-2024-07-18
+    display_name: GPT-4o mini (2024-07-18)
+    description: GPT-4o mini (2024-07-18) is a multimodal model with a context window of 128K tokens and improved handling of non-English text. ([blog](https://openai.com/index/gpt-4o-mini-advancing-cost-efficient-intelligence/))
+    creator_organization_name: OpenAI
+    access: limited
+    release_date: 2024-07-18
+    tags: [TEXT_MODEL_TAG, VISION_LANGUAGE_MODEL_TAG, OPENAI_CHATGPT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
   - name: openai/gpt-4-vision-preview
     # According to https://platform.openai.com/docs/models/gpt-4-turbo-and-gpt-4, this model has pointed gpt-4-1106-vision-preview.
     display_name: GPT-4V (1106 preview)
@@ -2074,30 +2221,30 @@ models:
   ## Codex Models
   # DEPRECATED: Codex models have been shut down on March 23 2023.
-  - name: openai/code-davinci-002 # DEPRECATED
+  - name: openai/code-davinci-002
     display_name: code-davinci-002
     description: Codex-style model that is designed for pure code-completion tasks ([docs](https://beta.openai.com/docs/models/codex)).
     creator_organization_name: OpenAI
     access: limited
     release_date: 2021-07-01 # TODO: Find correct date (this is for v1)
-    tags: [CODE_MODEL_TAG]
+    tags: [DEPRECATED_MODEL_TAG, CODE_MODEL_TAG]
-  - name: openai/code-davinci-001 # DEPRECATED
+  - name: openai/code-davinci-001
     display_name: code-davinci-001
     description: code-davinci-001 model
     creator_organization_name: OpenAI
     access: limited
     release_date: 2021-07-01 # Paper date
-    tags: [CODE_MODEL_TAG]
+    tags: [DEPRECATED_MODEL_TAG, CODE_MODEL_TAG]
-  - name: openai/code-cushman-001 # DEPRECATED
+  - name: openai/code-cushman-001
     display_name: code-cushman-001 (12B)
     description: Codex-style model that is a stronger, multilingual version of the Codex (12B) model in the [Codex paper](https://arxiv.org/pdf/2107.03374.pdf).
     creator_organization_name: OpenAI
     access: limited
     num_parameters: 12000000000
     release_date: 2021-07-01 # Paper date
-    tags: [CODE_MODEL_TAG]
+    tags: [DEPRECATED_MODEL_TAG, CODE_MODEL_TAG]
   ## Text Similarity Models
@@ -2107,41 +2254,41 @@ models:
   # DEPRECATED: Announced on July 06 2023 that first generation embeddings models
   #  will be shut down on January 04 2024.
-  - name: openai/text-similarity-davinci-001 # DEPRECATED
+  - name: openai/text-similarity-davinci-001
     display_name: text-similarity-davinci-001
     description: Embedding model that is designed for text similarity tasks ([docs](https://openai.com/blog/introducing-text-and-code-embeddings)).
     creator_organization_name: OpenAI
     access: limited
     num_parameters: 175000000000
     release_date: 2022-01-25 # Blog post date
-    tags: [TEXT_SIMILARITY_MODEL_TAG]
+    tags: [DEPRECATED_MODEL_TAG, TEXT_SIMILARITY_MODEL_TAG]
-  - name: openai/text-similarity-curie-001 # DEPRECATED
+  - name: openai/text-similarity-curie-001
     display_name: text-similarity-curie-001
     description: Embedding model that is designed for text similarity tasks ([docs](https://openai.com/blog/introducing-text-and-code-embeddings)).
     creator_organization_name: OpenAI
     access: limited
     num_parameters: 6700000000
     release_date: 2022-01-25 # Blog post date
-    tags: [TEXT_SIMILARITY_MODEL_TAG]
+    tags: [DEPRECATED_MODEL_TAG, TEXT_SIMILARITY_MODEL_TAG]
-  - name: openai/text-similarity-babbage-001 # DEPRECATED
+  - name: openai/text-similarity-babbage-001
     display_name: text-similarity-babbage-001
     description: Embedding model that is designed for text similarity tasks ([docs](https://openai.com/blog/introducing-text-and-code-embeddings)).
     creator_organization_name: OpenAI
     access: limited
     num_parameters: 1300000000
     release_date: 2022-01-25 # Blog post date
-    tags: [TEXT_SIMILARITY_MODEL_TAG]
+    tags: [DEPRECATED_MODEL_TAG, TEXT_SIMILARITY_MODEL_TAG]
-  - name: openai/text-similarity-ada-001 # DEPRECATED
+  - name: openai/text-similarity-ada-001
     display_name: text-similarity-ada-001
     description: Embedding model that is designed for text similarity tasks ([docs](https://openai.com/blog/introducing-text-and-code-embeddings)).
     creator_organization_name: OpenAI
     access: limited
     num_parameters: 350000000
     release_date: 2022-01-25 # Blog post date
-    tags: [TEXT_SIMILARITY_MODEL_TAG]
+    tags: [DEPRECATED_MODEL_TAG, TEXT_SIMILARITY_MODEL_TAG]
   - name: openai/text-embedding-ada-002
     display_name: text-embedding-ada-002
@@ -2197,6 +2344,34 @@ models:
     release_date: 2023-11-06
     tags: [TEXT_TO_IMAGE_MODEL_TAG]
+  # OpenThaiGPT
+  - name: openthaigpt/openthaigpt-1.0.0-7b-chat
+    display_name: OpenThaiGPT v1.0.0 (7B)
+    description: OpenThaiGPT v1.0.0 (7B) is a Thai language chat model based on Llama 2 that has been specifically fine-tuned for Thai instructions and enhanced by incorporating over 10,000 of the most commonly used Thai words into the dictionary. ([blog post](https://openthaigpt.aieat.or.th/openthaigpt-1.0.0-less-than-8-apr-2024-greater-than))
+    creator_organization_name: OpenThaiGPT
+    access: open
+    num_parameters: 7000000000
+    release_date: 2024-04-08
+    tags: [TEXT_MODEL_TAG, PARTIAL_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+  - name: openthaigpt/openthaigpt-1.0.0-13b-chat
+    display_name: OpenThaiGPT v1.0.0 (13B)
+    description: OpenThaiGPT v1.0.0 (13B) is a Thai language chat model based on Llama 2 that has been specifically fine-tuned for Thai instructions and enhanced by incorporating over 10,000 of the most commonly used Thai words into the dictionary. ([blog post](https://openthaigpt.aieat.or.th/openthaigpt-1.0.0-less-than-8-apr-2024-greater-than))
+    creator_organization_name: OpenThaiGPT
+    access: open
+    num_parameters: 13000000000
+    release_date: 2024-04-08
+    tags: [TEXT_MODEL_TAG, PARTIAL_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+  - name: openthaigpt/openthaigpt-1.0.0-70b-chat
+    display_name: OpenThaiGPT v1.0.0 (70B)
+    description: OpenThaiGPT v1.0.0 (70B) is a Thai language chat model based on Llama 2 that has been specifically fine-tuned for Thai instructions and enhanced by incorporating over 10,000 of the most commonly used Thai words into the dictionary. ([blog post](https://openthaigpt.aieat.or.th/openthaigpt-1.0.0-less-than-8-apr-2024-greater-than))
+    creator_organization_name: OpenThaiGPT
+    access: open
+    num_parameters: 70000000000
+    release_date: 2024-04-08
+    tags: [TEXT_MODEL_TAG, PARTIAL_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
   # Qwen
   - name: qwen/qwen-7b
@@ -2311,7 +2486,7 @@ models:
     access: open
     num_parameters: 7000000000
     release_date: 2024-04-04
-    tags: [TEXT_MODEL_TAG, PARTIAL_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+    tags: [TEXT_MODEL_TAG, PARTIAL_FUNCTIONALITY_TEXT_MODEL_TAG]
   - name: sail/sailor-7b-chat
     display_name: Sailor Chat (7B)
@@ -2329,7 +2504,7 @@ models:
     access: open
     num_parameters: 14000000000
     release_date: 2024-04-04
-    tags: [TEXT_MODEL_TAG, PARTIAL_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+    tags: [TEXT_MODEL_TAG, PARTIAL_FUNCTIONALITY_TEXT_MODEL_TAG]
   - name: sail/sailor-14b-chat
     display_name: Sailor Chat (14B)
@@ -2350,10 +2525,74 @@ models:
     release_date: 2022-03-25
     tags: [] # TODO: add tags
+  # SambaNova
+  - name: sambanova/sambalingo-thai-base
+    display_name: SambaLingo-Thai-Base
+    description: SambaLingo-Thai-Base is a pretrained bi-lingual Thai and English model that adapts Llama 2 (7B) to Thai by training on 38 billion tokens from the Thai split of the Cultura-X dataset. ([paper](https://arxiv.org/abs/2404.05829))
+    creator_organization_name: SambaLingo
+    access: open
+    num_parameters: 7000000000
+    release_date: 2024-04-08
+    tags: [TEXT_MODEL_TAG, PARTIAL_FUNCTIONALITY_TEXT_MODEL_TAG]
+  - name: sambanova/sambalingo-thai-chat
+    display_name: SambaLingo-Thai-Chat
+    description: SambaLingo-Thai-Chat is a chat model trained using direct preference optimization on SambaLingo-Thai-Base. SambaLingo-Thai-Base adapts Llama 2 (7B) to Thai by training on 38 billion tokens from the Thai split of the Cultura-X dataset. ([paper](https://arxiv.org/abs/2404.05829))
+    creator_organization_name: SambaLingo
+    access: open
+    num_parameters: 7000000000
+    release_date: 2024-04-08
+    tags: [TEXT_MODEL_TAG, PARTIAL_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+  - name: sambanova/sambalingo-thai-base-70b
+    display_name: SambaLingo-Thai-Base-70B
+    description: SambaLingo-Thai-Base-70B is a pretrained bi-lingual Thai and English model that adapts Llama 2 (70B) to Thai by training on 26 billion tokens from the Thai split of the Cultura-X dataset. ([paper](https://arxiv.org/abs/2404.05829))
+    creator_organization_name: SambaLingo
+    access: open
+    num_parameters: 70000000000
+    release_date: 2024-04-08
+    tags: [TEXT_MODEL_TAG, PARTIAL_FUNCTIONALITY_TEXT_MODEL_TAG]
+  - name: sambanova/sambalingo-thai-chat-70b
+    display_name: SambaLingo-Thai-Chat-70B
+    description: SambaLingo-Thai-Chat-70B is a chat model trained using direct preference optimization on SambaLingo-Thai-Base-70B. SambaLingo-Thai-Base-70B adapts Llama 2 (7B) to Thai by training on 26 billion tokens from the Thai split of the Cultura-X dataset. ([paper](https://arxiv.org/abs/2404.05829))
+    creator_organization_name: SambaLingo
+    access: open
+    num_parameters: 70000000000
+    release_date: 2024-04-08
+    tags: [TEXT_MODEL_TAG, PARTIAL_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
   # SCB10X
+  - name: scb10x/typhoon-7b
+    display_name: Typhoon (7B)
+    description: Typhoon (7B) is pretrained Thai large language model with 7 billion parameters based on Mistral 7B. ([paper](https://arxiv.org/abs/2312.13951))
+    creator_organization_name: SCB10X
+    access: open
+    num_parameters: 7000000000
+    release_date: 2023-12-21
+    tags: [TEXT_MODEL_TAG, PARTIAL_FUNCTIONALITY_TEXT_MODEL_TAG]
+  - name: scb10x/typhoon-v1.5-8b
+    display_name: Typhoon v1.5 (8B)
+    description: Typhoon v1.5 (8B) is a pretrained Thai large language model with 8 billion parameters based on Llama 3 8B. ([blog](https://blog.opentyphoon.ai/typhoon-1-5-release-a9364cb8e8d7))
+    creator_organization_name: SCB10X
+    access: open
+    num_parameters: 8000000000
+    release_date: 2024-05-08
+    tags: [TEXT_MODEL_TAG, PARTIAL_FUNCTIONALITY_TEXT_MODEL_TAG]
+  - name: scb10x/typhoon-v1.5-8b-instruct
+    display_name: Typhoon v1.5 Instruct (8B)
+    description: Typhoon v1.5 Instruct (8B) is a pretrained Thai large language model with 8 billion parameters based on Llama 3 8B. ([blog](https://blog.opentyphoon.ai/typhoon-1-5-release-a9364cb8e8d7))
+    creator_organization_name: SCB10X
+    access: open
+    num_parameters: 8000000000
+    release_date: 2024-05-08
+    tags: [TEXT_MODEL_TAG, PARTIAL_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
   - name: scb10x/typhoon-v1.5-72b
     display_name: Typhoon v1.5 (72B)
-    description: Typhoon v1.5 (72B) is pretrained Thai large language model with 72 billion parameters based on Qwen1.5-72B. ([blog](https://blog.opentyphoon.ai/typhoon-1-5-release-a9364cb8e8d7))
+    description: Typhoon v1.5 (72B) is a pretrained Thai large language model with 72 billion parameters based on Qwen1.5-72B. ([blog](https://blog.opentyphoon.ai/typhoon-1-5-release-a9364cb8e8d7))
     creator_organization_name: SCB10X
     access: open
     num_parameters: 72000000000
@@ -2362,13 +2601,50 @@ models:
   - name: scb10x/typhoon-v1.5-72b-instruct
     display_name: Typhoon v1.5 Instruct (72B)
-    description: Typhoon v1.5 Instruct (72B) is pretrained Thai large language model with 72 billion parameters based on Qwen1.5-72B. ([blog](https://blog.opentyphoon.ai/typhoon-1-5-release-a9364cb8e8d7))
+    description: Typhoon v1.5 Instruct (72B) is a pretrained Thai large language model with 72 billion parameters based on Qwen1.5-72B. ([blog](https://blog.opentyphoon.ai/typhoon-1-5-release-a9364cb8e8d7))
     creator_organization_name: SCB10X
     access: open
     num_parameters: 72000000000
     release_date: 2024-05-08
     tags: [TEXT_MODEL_TAG, PARTIAL_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+  - name: scb10x/llama-3-typhoon-v1.5x-8b-instruct
+    display_name: Typhoon 1.5X instruct (8B)
+    description: Llama-3-Typhoon-1.5X-8B-instruct is a 8 billion parameter instruct model designed for the Thai language based on Llama 3 Instruct. It utilizes the task-arithmetic model editing technique. ([blog](https://blog.opentyphoon.ai/typhoon-1-5x-our-experiment-designed-for-application-use-cases-7b85d9e9845c))
+    creator_organization_name: SCB10X
+    access: open
+    num_parameters: 8000000000
+    release_date: 2024-05-29
+    tags: [TEXT_MODEL_TAG, PARTIAL_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+  - name: scb10x/llama-3-typhoon-v1.5x-70b-instruct
+    display_name: Typhoon 1.5X instruct (70B)
+    description: Llama-3-Typhoon-1.5X-70B-instruct is a 70 billion parameter instruct model designed for the Thai language based on Llama 3 Instruct. It utilizes the task-arithmetic model editing technique. ([blog](https://blog.opentyphoon.ai/typhoon-1-5x-our-experiment-designed-for-application-use-cases-7b85d9e9845c))
+    creator_organization_name: SCB10X
+    access: open
+    num_parameters: 70000000000
+    release_date: 2024-05-29
+    tags: [TEXT_MODEL_TAG, PARTIAL_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+  # Alibaba DAMO Academy
+  - name: damo/seallm-7b-v2
+    display_name: SeaLLM v2 (7B)
+    description: SeaLLM v2 is a multilingual LLM for Southeast Asian (SEA) languages trained from Mistral (7B). ([website](https://damo-nlp-sg.github.io/SeaLLMs/))
+    creator_organization_name: Alibaba DAMO Academy
+    access: open
+    num_parameters: 7000000000
+    release_date: 2024-02-02
+    tags: [TEXT_MODEL_TAG, PARTIAL_FUNCTIONALITY_TEXT_MODEL_TAG]
+  - name: damo/seallm-7b-v2.5
+    display_name: SeaLLM v2.5 (7B)
+    description: SeaLLM is a multilingual LLM for Southeast Asian (SEA) languages trained from Gemma (7B). ([website](https://damo-nlp-sg.github.io/SeaLLMs/))
+    creator_organization_name: Alibaba DAMO Academy
+    access: open
+    num_parameters: 7000000000
+    release_date: 2024-04-12
+    tags: [TEXT_MODEL_TAG, PARTIAL_FUNCTIONALITY_TEXT_MODEL_TAG]
   # Snowflake
   - name: snowflake/snowflake-arctic-instruct
     display_name: Arctic Instruct
@@ -2538,7 +2814,7 @@ models:
     release_date: 2022-08-04
     # Inference with echo=True is not feasible -- in the prompt encoding phase, they use
     # bidirectional attention and do not perform predictions on them.
-    tags: [TEXT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, ABLATION_MODEL_TAG, NO_NEWLINES_TAG]
+    tags: [DEPRECATED_MODEL_TAG, TEXT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, ABLATION_MODEL_TAG, NO_NEWLINES_TAG]
   - name: tsinghua/codegeex # NOT SUPPORTED
     display_name: CodeGeeX (13B)
@@ -2573,7 +2849,6 @@ models:
     tags: [TEXT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG]
   - name: writer/palmyra-instruct-30
-    deprecated: true # Internal error
     display_name: InstructPalmyra (30B)
     description: InstructPalmyra (30B parameters) is trained using reinforcement learning techniques based on feedback from humans.
     creator_organization_name: Writer
@@ -2581,10 +2856,9 @@ models:
     num_parameters: 30000000000
     release_date: 2023-02-16
     # Does not support echo
-    tags: [TEXT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG]
+    tags: [DEPRECATED_MODEL_TAG, TEXT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG]
   - name: writer/palmyra-e
-    deprecated: true # Internal error
     display_name: Palmyra E (30B)
     description: Palmyra E (30B)
     creator_organization_name: Writer
@@ -2592,7 +2866,7 @@ models:
     num_parameters: 30000000000
     release_date: 2023-03-03
     # Does not support echo
-    tags: [TEXT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG]
+    tags: [DEPRECATED_MODEL_TAG, TEXT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG]
   - name: writer/silk-road
     display_name: Silk Road (35B)