Skip to content

Commit

Permalink
Merge pull request #387 from tryAGI/bot/update-together_202407220633
Browse files Browse the repository at this point in the history 
feat:Updated Together AI models
  • Loading branch information
@github-actions
github-actions[bot] authored Jul 22, 2024
2 parents 8aa60b3 + a90a649 commit 6397f57
Show file tree
Hide file tree
Showing 3 changed files with 141 additions and 56 deletions.
59 changes: 42 additions & 17 deletions src/Providers/TogetherAI/src/Predefined/AllModels.cs
View file
Original file line number Diff line number Diff line change
Expand Up @@ -215,6 +215,11 @@ public class Qwen15Chat4BModel(TogetherAiProvider provider) : TogetherAiModel(pr
public class MythomaxL213BModel(TogetherAiProvider provider) : TogetherAiModel(provider, TogetherAiModelIds.MythomaxL213B);


/// <inheritdoc cref="TogetherAiModelIds.MetaLlama370BReference"/>
/// <param name="provider">Open Router Provider Instance</param>
public class MetaLlama370BReferenceModel(TogetherAiProvider provider) : TogetherAiModel(provider, TogetherAiModelIds.MetaLlama370BReference);


/// <inheritdoc cref="TogetherAiModelIds.Mistral7BInstruct"/>
/// <param name="provider">Open Router Provider Instance</param>
public class Mistral7BInstructModel(TogetherAiProvider provider) : TogetherAiModel(provider, TogetherAiModelIds.Mistral7BInstruct);
Expand All @@ -230,11 +235,6 @@ public class Mistral7BInstructV02Model(TogetherAiProvider provider) : TogetherAi
public class OpenOrcaMistral7B8KModel(TogetherAiProvider provider) : TogetherAiModel(provider, TogetherAiModelIds.OpenOrcaMistral7B8K);


/// <inheritdoc cref="TogetherAiModelIds._01AiYiChat34B"/>
/// <param name="provider">Open Router Provider Instance</param>
public class _01AiYiChat34BModel(TogetherAiProvider provider) : TogetherAiModel(provider, TogetherAiModelIds._01AiYiChat34B);


/// <inheritdoc cref="TogetherAiModelIds.NousHermesLlama27B"/>
/// <param name="provider">Open Router Provider Instance</param>
public class NousHermesLlama27BModel(TogetherAiProvider provider) : TogetherAiModel(provider, TogetherAiModelIds.NousHermesLlama27B);
Expand Down Expand Up @@ -290,9 +290,44 @@ public class Qwen15Chat110BModel(TogetherAiProvider provider) : TogetherAiModel(
public class Llama2Chat13BModel(TogetherAiProvider provider) : TogetherAiModel(provider, TogetherAiModelIds.Llama2Chat13B);


/// <inheritdoc cref="TogetherAiModelIds.MetaLlama38BChat"/>
/// <inheritdoc cref="TogetherAiModelIds.Gemma2Instruct27B"/>
/// <param name="provider">Open Router Provider Instance</param>
public class MetaLlama38BChatModel(TogetherAiProvider provider) : TogetherAiModel(provider, TogetherAiModelIds.MetaLlama38BChat);
public class Gemma2Instruct27BModel(TogetherAiProvider provider) : TogetherAiModel(provider, TogetherAiModelIds.Gemma2Instruct27B);


/// <inheritdoc cref="TogetherAiModelIds._01AiYiChat34B"/>
/// <param name="provider">Open Router Provider Instance</param>
public class _01AiYiChat34BModel(TogetherAiProvider provider) : TogetherAiModel(provider, TogetherAiModelIds._01AiYiChat34B);


/// <inheritdoc cref="TogetherAiModelIds.MetaLlama370BInstructTurbo"/>
/// <param name="provider">Open Router Provider Instance</param>
public class MetaLlama370BInstructTurboModel(TogetherAiProvider provider) : TogetherAiModel(provider, TogetherAiModelIds.MetaLlama370BInstructTurbo);


/// <inheritdoc cref="TogetherAiModelIds.MetaLlama38BInstructTurbo"/>
/// <param name="provider">Open Router Provider Instance</param>
public class MetaLlama38BInstructTurboModel(TogetherAiProvider provider) : TogetherAiModel(provider, TogetherAiModelIds.MetaLlama38BInstructTurbo);


/// <inheritdoc cref="TogetherAiModelIds.MetaLlama370BInstructLite"/>
/// <param name="provider">Open Router Provider Instance</param>
public class MetaLlama370BInstructLiteModel(TogetherAiProvider provider) : TogetherAiModel(provider, TogetherAiModelIds.MetaLlama370BInstructLite);


/// <inheritdoc cref="TogetherAiModelIds.Gemma2Instruct9B"/>
/// <param name="provider">Open Router Provider Instance</param>
public class Gemma2Instruct9BModel(TogetherAiProvider provider) : TogetherAiModel(provider, TogetherAiModelIds.Gemma2Instruct9B);


/// <inheritdoc cref="TogetherAiModelIds.MetaLlama38BInstructLite"/>
/// <param name="provider">Open Router Provider Instance</param>
public class MetaLlama38BInstructLiteModel(TogetherAiProvider provider) : TogetherAiModel(provider, TogetherAiModelIds.MetaLlama38BInstructLite);


/// <inheritdoc cref="TogetherAiModelIds.MetaLlama38BReference"/>
/// <param name="provider">Open Router Provider Instance</param>
public class MetaLlama38BReferenceModel(TogetherAiProvider provider) : TogetherAiModel(provider, TogetherAiModelIds.MetaLlama38BReference);


/// <inheritdoc cref="TogetherAiModelIds.Mixtral8X7BInstructV01"/>
Expand Down Expand Up @@ -320,11 +355,6 @@ public class CodeLlama70BModel(TogetherAiProvider provider) : TogetherAiModel(pr
public class CodeLlamaPython7BModel(TogetherAiProvider provider) : TogetherAiModel(provider, TogetherAiModelIds.CodeLlamaPython7B);


/// <inheritdoc cref="TogetherAiModelIds.MetaLlama370BChat"/>
/// <param name="provider">Open Router Provider Instance</param>
public class MetaLlama370BChatModel(TogetherAiProvider provider) : TogetherAiModel(provider, TogetherAiModelIds.MetaLlama370BChat);


/// <inheritdoc cref="TogetherAiModelIds.ReservedDbrxInstruct"/>
/// <param name="provider">Open Router Provider Instance</param>
public class ReservedDbrxInstructModel(TogetherAiProvider provider) : TogetherAiModel(provider, TogetherAiModelIds.ReservedDbrxInstruct);
Expand All @@ -350,11 +380,6 @@ public class MetaLlama38BInstructModel(TogetherAiProvider provider) : TogetherAi
public class MetaLlama370BInstructModel(TogetherAiProvider provider) : TogetherAiModel(provider, TogetherAiModelIds.MetaLlama370BInstruct);


/// <inheritdoc cref="TogetherAiModelIds.Gemma2Instruct9B"/>
/// <param name="provider">Open Router Provider Instance</param>
public class Gemma2Instruct9BModel(TogetherAiProvider provider) : TogetherAiModel(provider, TogetherAiModelIds.Gemma2Instruct9B);


/// <inheritdoc cref="TogetherAiModelIds.Hermes2ThetaLlama370B"/>
/// <param name="provider">Open Router Provider Instance</param>
public class Hermes2ThetaLlama370BModel(TogetherAiProvider provider) : TogetherAiModel(provider, TogetherAiModelIds.Hermes2ThetaLlama370B);
Expand Down
125 changes: 90 additions & 35 deletions src/Providers/TogetherAI/src/TogetherAiModelIds.cs
View file
Original file line number Diff line number Diff line change
Expand Up @@ -479,6 +479,17 @@ public enum TogetherAiModelIds
/// </summary>
MythomaxL213B,

/// <summary>
/// Name: Meta Llama 3 70B Reference <br/>
/// Organization: Meta <br/>
/// Context Length: 8192 <br/>
/// Prompt Cost: $0.9/MTok <br/>
/// Completion Cost: $0.9/MTok <br/>
/// Description: Llama 3 is an auto-regressive language model that uses an optimized transformer architecture. The tuned versions use supervised fine-tuning (SFT) and reinforcement learning with human feedback (RLHF) to align with human preferences for helpfulness and safety. <br/>
/// HuggingFace Url: <a href="https://huggingface.co/meta-llama/Llama-3-70b-chat-hf">https://huggingface.co/meta-llama/Llama-3-70b-chat-hf</a>
/// </summary>
MetaLlama370BReference,

/// <summary>
/// Name: Mistral (7B) Instruct <br/>
/// Organization: mistralai <br/>
Expand Down Expand Up @@ -512,17 +523,6 @@ public enum TogetherAiModelIds
/// </summary>
OpenOrcaMistral7B8K,

/// <summary>
/// Name: 01-ai Yi Chat (34B) <br/>
/// Organization: 01.AI <br/>
/// Context Length: 4096 <br/>
/// Prompt Cost: $0.8/MTok <br/>
/// Completion Cost: $0.8/MTok <br/>
/// Description: The Yi series models are large language models trained from scratch by developers at 01.AI <br/>
/// HuggingFace Url: <a href="https://huggingface.co/zero-one-ai/Yi-34B-Chat">https://huggingface.co/zero-one-ai/Yi-34B-Chat</a>
/// </summary>
_01AiYiChat34B,

/// <summary>
/// Name: Nous Hermes LLaMA-2 (7B) <br/>
/// Organization: NousResearch <br/>
Expand Down Expand Up @@ -645,15 +645,92 @@ public enum TogetherAiModelIds
Llama2Chat13B,

/// <summary>
/// Name: Meta Llama 3 8B Chat <br/>
/// Name: Gemma-2 Instruct (27B) <br/>
/// Organization: Google <br/>
/// Context Length: 8192 <br/>
/// Prompt Cost: $0.8/MTok <br/>
/// Completion Cost: $0.8/MTok <br/>
/// 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. <br/>
/// HuggingFace Url: <a href="https://huggingface.co/google/gemma-2-27b-it">https://huggingface.co/google/gemma-2-27b-it</a>
/// </summary>
Gemma2Instruct27B,

/// <summary>
/// Name: 01-ai Yi Chat (34B) <br/>
/// Organization: 01.AI <br/>
/// Context Length: 4096 <br/>
/// Prompt Cost: $0.8/MTok <br/>
/// Completion Cost: $0.8/MTok <br/>
/// Description: The Yi series models are large language models trained from scratch by developers at 01.AI <br/>
/// HuggingFace Url: <a href="https://huggingface.co/zero-one-ai/Yi-34B-Chat">https://huggingface.co/zero-one-ai/Yi-34B-Chat</a>
/// </summary>
_01AiYiChat34B,

/// <summary>
/// Name: Meta Llama 3 70B Instruct Turbo <br/>
/// Organization: Meta <br/>
/// Context Length: 8192 <br/>
/// Prompt Cost: $0.88/MTok <br/>
/// Completion Cost: $0.88/MTok <br/>
/// Description: Llama 3 is an auto-regressive language model that uses an optimized transformer architecture. The tuned versions use supervised fine-tuning (SFT) and reinforcement learning with human feedback (RLHF) to align with human preferences for helpfulness and safety. <br/>
/// HuggingFace Url: <a href="https://huggingface.co/meta-llama/Meta-Llama-3-70B-Instruct-Turbo">https://huggingface.co/meta-llama/Meta-Llama-3-70B-Instruct-Turbo</a>
/// </summary>
MetaLlama370BInstructTurbo,

/// <summary>
/// Name: Meta Llama 3 8B Instruct Turbo <br/>
/// Organization: Meta <br/>
/// Context Length: 8192 <br/>
/// Prompt Cost: $0.18/MTok <br/>
/// Completion Cost: $0.18/MTok <br/>
/// Description: Llama 3 is an auto-regressive language model that uses an optimized transformer architecture. The tuned versions use supervised fine-tuning (SFT) and reinforcement learning with human feedback (RLHF) to align with human preferences for helpfulness and safety. <br/>
/// HuggingFace Url: <a href="https://huggingface.co/meta-llama/Meta-Llama-3-8B-Instruct-Turbo">https://huggingface.co/meta-llama/Meta-Llama-3-8B-Instruct-Turbo</a>
/// </summary>
MetaLlama38BInstructTurbo,

/// <summary>
/// Name: Meta Llama 3 70B Instruct Lite <br/>
/// Organization: Meta <br/>
/// Context Length: 8192 <br/>
/// Prompt Cost: $0.54/MTok <br/>
/// Completion Cost: $0.54/MTok <br/>
/// Description: Llama 3 is an auto-regressive language model that uses an optimized transformer architecture. The tuned versions use supervised fine-tuning (SFT) and reinforcement learning with human feedback (RLHF) to align with human preferences for helpfulness and safety. <br/>
/// HuggingFace Url: <a href="https://huggingface.co/meta-llama/Meta-Llama-3-70B-Instruct-Lite">https://huggingface.co/meta-llama/Meta-Llama-3-70B-Instruct-Lite</a>
/// </summary>
MetaLlama370BInstructLite,

/// <summary>
/// Name: Gemma-2 Instruct (9B) <br/>
/// Organization: google <br/>
/// Context Length: 8192 <br/>
/// Prompt Cost: $0.3/MTok <br/>
/// Completion Cost: $0.3/MTok <br/>
/// 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. <br/>
/// HuggingFace Url: <a href="https://huggingface.co/google/gemma-2-9b-it">https://huggingface.co/google/gemma-2-9b-it</a>
/// </summary>
Gemma2Instruct9B,

/// <summary>
/// Name: Meta Llama 3 8B Instruct Lite <br/>
/// Organization: Meta <br/>
/// Context Length: 8192 <br/>
/// Prompt Cost: $0.1/MTok <br/>
/// Completion Cost: $0.1/MTok <br/>
/// Description: Llama 3 is an auto-regressive language model that uses an optimized transformer architecture. The tuned versions use supervised fine-tuning (SFT) and reinforcement learning with human feedback (RLHF) to align with human preferences for helpfulness and safety. <br/>
/// HuggingFace Url: <a href="https://huggingface.co/meta-llama/Meta-Llama-3-8B-Instruct-Lite">https://huggingface.co/meta-llama/Meta-Llama-3-8B-Instruct-Lite</a>
/// </summary>
MetaLlama38BInstructLite,

/// <summary>
/// Name: Meta Llama 3 8B Reference <br/>
/// Organization: Meta <br/>
/// Context Length: 8192 <br/>
/// Prompt Cost: $0.2/MTok <br/>
/// Completion Cost: $0.2/MTok <br/>
/// Description: Llama 3 is an auto-regressive language model that uses an optimized transformer architecture. The tuned versions use supervised fine-tuning (SFT) and reinforcement learning with human feedback (RLHF) to align with human preferences for helpfulness and safety. <br/>
/// HuggingFace Url: <a href="https://huggingface.co/meta-llama/Llama-3-8b-chat-hf">https://huggingface.co/meta-llama/Llama-3-8b-chat-hf</a>
/// </summary>
MetaLlama38BChat,
MetaLlama38BReference,

/// <summary>
/// Name: Mixtral-8x7B Instruct v0.1 <br/>
Expand Down Expand Up @@ -710,17 +787,6 @@ public enum TogetherAiModelIds
/// </summary>
CodeLlamaPython7B,

/// <summary>
/// Name: Meta Llama 3 70B Chat <br/>
/// Organization: Meta <br/>
/// Context Length: 8192 <br/>
/// Prompt Cost: $0.9/MTok <br/>
/// Completion Cost: $0.9/MTok <br/>
/// Description: Llama 3 is an auto-regressive language model that uses an optimized transformer architecture. The tuned versions use supervised fine-tuning (SFT) and reinforcement learning with human feedback (RLHF) to align with human preferences for helpfulness and safety. <br/>
/// HuggingFace Url: <a href="https://huggingface.co/meta-llama/Llama-3-70b-chat-hf">https://huggingface.co/meta-llama/Llama-3-70b-chat-hf</a>
/// </summary>
MetaLlama370BChat,

/// <summary>
/// Name: Reserved - DBRX Instruct <br/>
/// Organization: Databricks <br/>
Expand Down Expand Up @@ -776,17 +842,6 @@ public enum TogetherAiModelIds
/// </summary>
MetaLlama370BInstruct,

/// <summary>
/// Name: Gemma-2 Instruct (9B) <br/>
/// Organization: google <br/>
/// Context Length: 8192 <br/>
/// Prompt Cost: $0.3/MTok <br/>
/// Completion Cost: $0.3/MTok <br/>
/// 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. <br/>
/// HuggingFace Url: <a href="https://huggingface.co/google/gemma-2-9b-it">https://huggingface.co/google/gemma-2-9b-it</a>
/// </summary>
Gemma2Instruct9B,

/// <summary>
/// Name: Hermes 2 Theta Llama-3 70B <br/>
/// Organization: NousResearch <br/>
Expand Down
13 changes: 9 additions & 4 deletions src/Providers/TogetherAI/src/TogetherAiModelProvider.cs
View file
Original file line number Diff line number Diff line change
Expand Up @@ -52,10 +52,10 @@ public static class TogetherAiModelProvider
{ TogetherAiModelIds.OlmoInstruct7B, ToMetadata("allenai/OLMo-7B-Instruct",2048,2.0000000000000002E-07,2.0000000000000002E-07)},
{ TogetherAiModelIds.Qwen15Chat4B, ToMetadata("Qwen/Qwen1.5-4B-Chat",32768,1.0000000000000001E-07,1.0000000000000001E-07)},
{ TogetherAiModelIds.MythomaxL213B, ToMetadata("Gryphe/MythoMax-L2-13b",4096,3E-07,3E-07)},
{ TogetherAiModelIds.MetaLlama370BReference, ToMetadata("meta-llama/Llama-3-70b-chat-hf",8192,9.000000000000001E-07,9.000000000000001E-07)},
{ TogetherAiModelIds.Mistral7BInstruct, ToMetadata("mistralai/Mistral-7B-Instruct-v0.1",4096,2.0000000000000002E-07,2.0000000000000002E-07)},
{ TogetherAiModelIds.Mistral7BInstructV02, ToMetadata("mistralai/Mistral-7B-Instruct-v0.2",32768,2.0000000000000002E-07,2.0000000000000002E-07)},
{ TogetherAiModelIds.OpenOrcaMistral7B8K, ToMetadata("Open-Orca/Mistral-7B-OpenOrca",8192,2.0000000000000002E-07,2.0000000000000002E-07)},
{ TogetherAiModelIds._01AiYiChat34B, ToMetadata("zero-one-ai/Yi-34B-Chat",4096,8.000000000000001E-07,8.000000000000001E-07)},
{ TogetherAiModelIds.NousHermesLlama27B, ToMetadata("NousResearch/Nous-Hermes-llama-2-7b",4096,2.0000000000000002E-07,2.0000000000000002E-07)},
{ TogetherAiModelIds.Qwen15Chat32B, ToMetadata("Qwen/Qwen1.5-32B-Chat",32768,8.000000000000001E-07,8.000000000000001E-07)},
{ TogetherAiModelIds.Qwen2Instruct72B, ToMetadata("Qwen/Qwen2-72B-Instruct",32768,9.000000000000001E-07,9.000000000000001E-07)},
Expand All @@ -67,19 +67,24 @@ public static class TogetherAiModelProvider
{ TogetherAiModelIds.Mistral7BInstructV03, ToMetadata("mistralai/Mistral-7B-Instruct-v0.3",32768,2.0000000000000002E-07,2.0000000000000002E-07)},
{ TogetherAiModelIds.Qwen15Chat110B, ToMetadata("Qwen/Qwen1.5-110B-Chat",32768,1.8000000000000001E-06,1.8000000000000001E-06)},
{ TogetherAiModelIds.Llama2Chat13B, ToMetadata("meta-llama/Llama-2-13b-chat-hf",4096,2.2E-07,2.2E-07)},
{ TogetherAiModelIds.MetaLlama38BChat, ToMetadata("meta-llama/Llama-3-8b-chat-hf",8192,2.0000000000000002E-07,2.0000000000000002E-07)},
{ TogetherAiModelIds.Gemma2Instruct27B, ToMetadata("google/gemma-2-27b-it",8192,8.000000000000001E-07,8.000000000000001E-07)},
{ TogetherAiModelIds._01AiYiChat34B, ToMetadata("zero-one-ai/Yi-34B-Chat",4096,8.000000000000001E-07,8.000000000000001E-07)},
{ TogetherAiModelIds.MetaLlama370BInstructTurbo, ToMetadata("meta-llama/Meta-Llama-3-70B-Instruct-Turbo",8192,8.8E-07,8.8E-07)},
{ TogetherAiModelIds.MetaLlama38BInstructTurbo, ToMetadata("meta-llama/Meta-Llama-3-8B-Instruct-Turbo",8192,1.8E-07,1.8E-07)},
{ TogetherAiModelIds.MetaLlama370BInstructLite, ToMetadata("meta-llama/Meta-Llama-3-70B-Instruct-Lite",8192,5.4E-07,5.4E-07)},
{ TogetherAiModelIds.Gemma2Instruct9B, ToMetadata("google/gemma-2-9b-it",8192,3E-07,3E-07)},
{ TogetherAiModelIds.MetaLlama38BInstructLite, ToMetadata("meta-llama/Meta-Llama-3-8B-Instruct-Lite",8192,1.0000000000000001E-07,1.0000000000000001E-07)},
{ TogetherAiModelIds.MetaLlama38BReference, ToMetadata("meta-llama/Llama-3-8b-chat-hf",8192,2.0000000000000002E-07,2.0000000000000002E-07)},
{ TogetherAiModelIds.Mixtral8X7BInstructV01, ToMetadata("mistralai/Mixtral-8x7B-Instruct-v0.1",32768,6E-07,6E-07)},
{ TogetherAiModelIds.Llama2Chat7B, ToMetadata("meta-llama/Llama-2-7b-chat-hf",4096,2.0000000000000002E-07,2.0000000000000002E-07)},
{ TogetherAiModelIds.Llama2Chat70B, ToMetadata("meta-llama/Llama-2-70b-chat-hf",4096,9.000000000000001E-07,9.000000000000001E-07)},
{ TogetherAiModelIds.CodeLlama70B, ToMetadata("codellama/CodeLlama-70b-hf",16384,9.000000000000001E-07,9.000000000000001E-07)},
{ TogetherAiModelIds.CodeLlamaPython7B, ToMetadata("codellama/CodeLlama-7b-Python-hf",16384,2.0000000000000002E-07,2.0000000000000002E-07)},
{ TogetherAiModelIds.MetaLlama370BChat, ToMetadata("meta-llama/Llama-3-70b-chat-hf",8192,9.000000000000001E-07,9.000000000000001E-07)},
{ TogetherAiModelIds.ReservedDbrxInstruct, ToMetadata("medaltv/dbrx-instruct",0,1.2E-06,1.2E-06)},
{ TogetherAiModelIds.DbrxInstruct, ToMetadata("databricks/dbrx-instruct",32768,1.2E-06,1.2E-06)},
{ TogetherAiModelIds.UpstageSolarInstructV111BInt4, ToMetadata("togethercomputer/SOLAR-10.7B-Instruct-v1.0-int4",4096,3E-07,3E-07)},
{ TogetherAiModelIds.MetaLlama38BInstruct, ToMetadata("meta-llama/Meta-Llama-3-8B-Instruct",8192,2.0000000000000002E-07,2.0000000000000002E-07)},
{ TogetherAiModelIds.MetaLlama370BInstruct, ToMetadata("meta-llama/Meta-Llama-3-70B-Instruct",8192,9.000000000000001E-07,9.000000000000001E-07)},
{ TogetherAiModelIds.Gemma2Instruct9B, ToMetadata("google/gemma-2-9b-it",8192,3E-07,3E-07)},
{ TogetherAiModelIds.Hermes2ThetaLlama370B, ToMetadata("NousResearch/Hermes-2-Theta-Llama-3-70B",8192,0,0)},
{ TogetherAiModelIds.Llama370BInstructGradient1048K, ToMetadata("gradientai/Llama-3-70B-Instruct-Gradient-1048k",1048576,0,0)},

Expand Down

0 comments on commit 6397f57

Please sign in to comment.