diff --git a/api/core/model_runtime/model_providers/openrouter/llm/deepseek-chat.yaml b/api/core/model_runtime/model_providers/openrouter/llm/deepseek-chat.yaml index 7a1dea69503..6743bfcad66 100644 --- a/api/core/model_runtime/model_providers/openrouter/llm/deepseek-chat.yaml +++ b/api/core/model_runtime/model_providers/openrouter/llm/deepseek-chat.yaml @@ -35,6 +35,15 @@ parameter_rules: help: zh_Hans: 控制生成结果的随机性。数值越小，随机性越弱；数值越大，随机性越强。一般而言，top_p 和 temperature 两个参数选择一个进行调整即可。 en_US: Control the randomness of generated results. The smaller the value, the weaker the randomness; the larger the value, the stronger the randomness. Generally speaking, you can adjust one of the two parameters top_p and temperature. + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false - name: frequency_penalty use_template: frequency_penalty default: 0 diff --git a/api/core/model_runtime/model_providers/openrouter/llm/deepseek-coder.yaml b/api/core/model_runtime/model_providers/openrouter/llm/deepseek-coder.yaml index c05f4769b83..375a4d2d52e 100644 --- a/api/core/model_runtime/model_providers/openrouter/llm/deepseek-coder.yaml +++ b/api/core/model_runtime/model_providers/openrouter/llm/deepseek-coder.yaml @@ -18,6 +18,15 @@ parameter_rules: min: 0 max: 1 default: 1 + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false - name: max_tokens use_template: max_tokens min: 1 diff --git a/api/core/model_runtime/model_providers/openrouter/llm/gpt-3.5-turbo.yaml b/api/core/model_runtime/model_providers/openrouter/llm/gpt-3.5-turbo.yaml index 186c1cc6636..621ecf065e9 100644 --- a/api/core/model_runtime/model_providers/openrouter/llm/gpt-3.5-turbo.yaml +++ b/api/core/model_runtime/model_providers/openrouter/llm/gpt-3.5-turbo.yaml @@ -14,6 +14,15 @@ parameter_rules: use_template: temperature - name: top_p use_template: top_p + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false - name: presence_penalty use_template: presence_penalty - name: frequency_penalty diff --git a/api/core/model_runtime/model_providers/openrouter/llm/gpt-4-32k.yaml b/api/core/model_runtime/model_providers/openrouter/llm/gpt-4-32k.yaml index 8c2989b3000..887e6d60f9c 100644 --- a/api/core/model_runtime/model_providers/openrouter/llm/gpt-4-32k.yaml +++ b/api/core/model_runtime/model_providers/openrouter/llm/gpt-4-32k.yaml @@ -14,6 +14,15 @@ parameter_rules: use_template: temperature - name: top_p use_template: top_p + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false - name: presence_penalty use_template: presence_penalty - name: frequency_penalty diff --git a/api/core/model_runtime/model_providers/openrouter/llm/gpt-4.yaml b/api/core/model_runtime/model_providers/openrouter/llm/gpt-4.yaml index ef19d4f6f0d..66d1f9ae67e 100644 --- a/api/core/model_runtime/model_providers/openrouter/llm/gpt-4.yaml +++ b/api/core/model_runtime/model_providers/openrouter/llm/gpt-4.yaml @@ -14,6 +14,15 @@ parameter_rules: use_template: temperature - name: top_p use_template: top_p + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false - name: presence_penalty use_template: presence_penalty - name: frequency_penalty diff --git a/api/core/model_runtime/model_providers/openrouter/llm/gpt-4o-2024-08-06.yaml b/api/core/model_runtime/model_providers/openrouter/llm/gpt-4o-2024-08-06.yaml index 0be325f55bd..695cc3eedf5 100644 --- a/api/core/model_runtime/model_providers/openrouter/llm/gpt-4o-2024-08-06.yaml +++ b/api/core/model_runtime/model_providers/openrouter/llm/gpt-4o-2024-08-06.yaml @@ -16,6 +16,15 @@ parameter_rules: use_template: temperature - name: top_p use_template: top_p + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false - name: presence_penalty use_template: presence_penalty - name: frequency_penalty diff --git a/api/core/model_runtime/model_providers/openrouter/llm/gpt-4o-mini.yaml b/api/core/model_runtime/model_providers/openrouter/llm/gpt-4o-mini.yaml index 3b1d95643d2..e1e5889085f 100644 --- a/api/core/model_runtime/model_providers/openrouter/llm/gpt-4o-mini.yaml +++ b/api/core/model_runtime/model_providers/openrouter/llm/gpt-4o-mini.yaml @@ -15,6 +15,15 @@ parameter_rules: use_template: temperature - name: top_p use_template: top_p + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false - name: presence_penalty use_template: presence_penalty - name: frequency_penalty diff --git a/api/core/model_runtime/model_providers/openrouter/llm/gpt-4o.yaml b/api/core/model_runtime/model_providers/openrouter/llm/gpt-4o.yaml index a8c97efdd64..560bf9d7d05 100644 --- a/api/core/model_runtime/model_providers/openrouter/llm/gpt-4o.yaml +++ b/api/core/model_runtime/model_providers/openrouter/llm/gpt-4o.yaml @@ -15,6 +15,15 @@ parameter_rules: use_template: temperature - name: top_p use_template: top_p + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false - name: presence_penalty use_template: presence_penalty - name: frequency_penalty diff --git a/api/core/model_runtime/model_providers/openrouter/llm/llama-3-70b-instruct.yaml b/api/core/model_runtime/model_providers/openrouter/llm/llama-3-70b-instruct.yaml index b91c39e729e..04a4a90c6d6 100644 --- a/api/core/model_runtime/model_providers/openrouter/llm/llama-3-70b-instruct.yaml +++ b/api/core/model_runtime/model_providers/openrouter/llm/llama-3-70b-instruct.yaml @@ -10,6 +10,15 @@ parameter_rules: use_template: temperature - name: top_p use_template: top_p + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false - name: max_tokens use_template: max_tokens required: true diff --git a/api/core/model_runtime/model_providers/openrouter/llm/llama-3-8b-instruct.yaml b/api/core/model_runtime/model_providers/openrouter/llm/llama-3-8b-instruct.yaml index 84b2c7fac2c..066949d431e 100644 --- a/api/core/model_runtime/model_providers/openrouter/llm/llama-3-8b-instruct.yaml +++ b/api/core/model_runtime/model_providers/openrouter/llm/llama-3-8b-instruct.yaml @@ -10,6 +10,15 @@ parameter_rules: use_template: temperature - name: top_p use_template: top_p + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false - name: max_tokens use_template: max_tokens required: true diff --git a/api/core/model_runtime/model_providers/openrouter/llm/llama-3.1-405b-instruct.yaml b/api/core/model_runtime/model_providers/openrouter/llm/llama-3.1-405b-instruct.yaml index a489ce1b5ad..0cd89dea716 100644 --- a/api/core/model_runtime/model_providers/openrouter/llm/llama-3.1-405b-instruct.yaml +++ b/api/core/model_runtime/model_providers/openrouter/llm/llama-3.1-405b-instruct.yaml @@ -10,6 +10,15 @@ parameter_rules: use_template: temperature - name: top_p use_template: top_p + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false - name: max_tokens use_template: max_tokens required: true diff --git a/api/core/model_runtime/model_providers/openrouter/llm/llama-3.1-70b-instruct.yaml b/api/core/model_runtime/model_providers/openrouter/llm/llama-3.1-70b-instruct.yaml index 12037411b10..768ab5ecbb6 100644 --- a/api/core/model_runtime/model_providers/openrouter/llm/llama-3.1-70b-instruct.yaml +++ b/api/core/model_runtime/model_providers/openrouter/llm/llama-3.1-70b-instruct.yaml @@ -10,6 +10,15 @@ parameter_rules: use_template: temperature - name: top_p use_template: top_p + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false - name: max_tokens use_template: max_tokens required: true diff --git a/api/core/model_runtime/model_providers/openrouter/llm/llama-3.1-8b-instruct.yaml b/api/core/model_runtime/model_providers/openrouter/llm/llama-3.1-8b-instruct.yaml index 6f06493f293..67b6b82b5d7 100644 --- a/api/core/model_runtime/model_providers/openrouter/llm/llama-3.1-8b-instruct.yaml +++ b/api/core/model_runtime/model_providers/openrouter/llm/llama-3.1-8b-instruct.yaml @@ -10,6 +10,15 @@ parameter_rules: use_template: temperature - name: top_p use_template: top_p + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false - name: max_tokens use_template: max_tokens required: true diff --git a/api/core/model_runtime/model_providers/openrouter/llm/mistral-7b-instruct.yaml b/api/core/model_runtime/model_providers/openrouter/llm/mistral-7b-instruct.yaml index 012dfc55ce1..d08c016e954 100644 --- a/api/core/model_runtime/model_providers/openrouter/llm/mistral-7b-instruct.yaml +++ b/api/core/model_runtime/model_providers/openrouter/llm/mistral-7b-instruct.yaml @@ -18,6 +18,15 @@ parameter_rules: default: 1 min: 0 max: 1 + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false - name: max_tokens use_template: max_tokens default: 1024 diff --git a/api/core/model_runtime/model_providers/openrouter/llm/mixtral-8x22b-instruct.yaml b/api/core/model_runtime/model_providers/openrouter/llm/mixtral-8x22b-instruct.yaml index f4eb4e45d95..e3af0e64d8f 100644 --- a/api/core/model_runtime/model_providers/openrouter/llm/mixtral-8x22b-instruct.yaml +++ b/api/core/model_runtime/model_providers/openrouter/llm/mixtral-8x22b-instruct.yaml @@ -18,6 +18,15 @@ parameter_rules: default: 1 min: 0 max: 1 + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false - name: max_tokens use_template: max_tokens default: 1024 diff --git a/api/core/model_runtime/model_providers/openrouter/llm/mixtral-8x7b-instruct.yaml b/api/core/model_runtime/model_providers/openrouter/llm/mixtral-8x7b-instruct.yaml index 7871e1f7a05..095ea5a8584 100644 --- a/api/core/model_runtime/model_providers/openrouter/llm/mixtral-8x7b-instruct.yaml +++ b/api/core/model_runtime/model_providers/openrouter/llm/mixtral-8x7b-instruct.yaml @@ -19,6 +19,15 @@ parameter_rules: default: 1 min: 0 max: 1 + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false - name: max_tokens use_template: max_tokens default: 1024 diff --git a/api/core/model_runtime/model_providers/openrouter/llm/o1-mini.yaml b/api/core/model_runtime/model_providers/openrouter/llm/o1-mini.yaml index 85a918ff5e0..f4202ee814e 100644 --- a/api/core/model_runtime/model_providers/openrouter/llm/o1-mini.yaml +++ b/api/core/model_runtime/model_providers/openrouter/llm/o1-mini.yaml @@ -12,6 +12,15 @@ parameter_rules: use_template: temperature - name: top_p use_template: top_p + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false - name: presence_penalty use_template: presence_penalty - name: frequency_penalty diff --git a/api/core/model_runtime/model_providers/openrouter/llm/o1-preview.yaml b/api/core/model_runtime/model_providers/openrouter/llm/o1-preview.yaml index 74b0a511bed..1281b842862 100644 --- a/api/core/model_runtime/model_providers/openrouter/llm/o1-preview.yaml +++ b/api/core/model_runtime/model_providers/openrouter/llm/o1-preview.yaml @@ -12,6 +12,15 @@ parameter_rules: use_template: temperature - name: top_p use_template: top_p + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false - name: presence_penalty use_template: presence_penalty - name: frequency_penalty diff --git a/api/core/model_runtime/model_providers/openrouter/llm/qwen2-72b-instruct.yaml b/api/core/model_runtime/model_providers/openrouter/llm/qwen2-72b-instruct.yaml index 7b75fcb0c98..b6058138d30 100644 --- a/api/core/model_runtime/model_providers/openrouter/llm/qwen2-72b-instruct.yaml +++ b/api/core/model_runtime/model_providers/openrouter/llm/qwen2-72b-instruct.yaml @@ -21,6 +21,15 @@ parameter_rules: en_US: Specifies the upper limit on the length of generated results. If the generated results are truncated, you can increase this parameter. - name: top_p use_template: top_p + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false - name: frequency_penalty use_template: frequency_penalty pricing: diff --git a/api/core/model_runtime/model_providers/openrouter/llm/qwen2.5-72b-instruct.yaml b/api/core/model_runtime/model_providers/openrouter/llm/qwen2.5-72b-instruct.yaml index f141a40a00c..5392b111689 100644 --- a/api/core/model_runtime/model_providers/openrouter/llm/qwen2.5-72b-instruct.yaml +++ b/api/core/model_runtime/model_providers/openrouter/llm/qwen2.5-72b-instruct.yaml @@ -21,6 +21,15 @@ parameter_rules: en_US: Specifies the upper limit on the length of generated results. If the generated results are truncated, you can increase this parameter. - name: top_p use_template: top_p + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false - name: frequency_penalty use_template: frequency_penalty pricing: diff --git a/api/core/model_runtime/model_providers/siliconflow/llm/deepdeek-coder-v2-instruct.yaml b/api/core/model_runtime/model_providers/siliconflow/llm/deepdeek-coder-v2-instruct.yaml index d4431179e56..d5f23776ea2 100644 --- a/api/core/model_runtime/model_providers/siliconflow/llm/deepdeek-coder-v2-instruct.yaml +++ b/api/core/model_runtime/model_providers/siliconflow/llm/deepdeek-coder-v2-instruct.yaml @@ -21,6 +21,15 @@ parameter_rules: en_US: Specifies the upper limit on the length of generated results. If the generated results are truncated, you can increase this parameter. - name: top_p use_template: top_p + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false - name: frequency_penalty use_template: frequency_penalty pricing: diff --git a/api/core/model_runtime/model_providers/siliconflow/llm/deepseek-v2-chat.yaml b/api/core/model_runtime/model_providers/siliconflow/llm/deepseek-v2-chat.yaml index caa6508b5ed..7aa684ef381 100644 --- a/api/core/model_runtime/model_providers/siliconflow/llm/deepseek-v2-chat.yaml +++ b/api/core/model_runtime/model_providers/siliconflow/llm/deepseek-v2-chat.yaml @@ -21,6 +21,15 @@ parameter_rules: en_US: Specifies the upper limit on the length of generated results. If the generated results are truncated, you can increase this parameter. - name: top_p use_template: top_p + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false - name: frequency_penalty use_template: frequency_penalty pricing: diff --git a/api/core/model_runtime/model_providers/siliconflow/llm/deepseek-v2.5.yaml b/api/core/model_runtime/model_providers/siliconflow/llm/deepseek-v2.5.yaml index 1c8e15ae52a..b30fa3e2d15 100644 --- a/api/core/model_runtime/model_providers/siliconflow/llm/deepseek-v2.5.yaml +++ b/api/core/model_runtime/model_providers/siliconflow/llm/deepseek-v2.5.yaml @@ -21,6 +21,15 @@ parameter_rules: en_US: Specifies the upper limit on the length of generated results. If the generated results are truncated, you can increase this parameter. - name: top_p use_template: top_p + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false - name: frequency_penalty use_template: frequency_penalty pricing: diff --git a/api/core/model_runtime/model_providers/siliconflow/llm/gemma-2-27b-it.yaml b/api/core/model_runtime/model_providers/siliconflow/llm/gemma-2-27b-it.yaml index 2840e3dcf4b..f2a1f64bfb5 100644 --- a/api/core/model_runtime/model_providers/siliconflow/llm/gemma-2-27b-it.yaml +++ b/api/core/model_runtime/model_providers/siliconflow/llm/gemma-2-27b-it.yaml @@ -21,6 +21,15 @@ parameter_rules: en_US: Specifies the upper limit on the length of generated results. If the generated results are truncated, you can increase this parameter. - name: top_p use_template: top_p + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false - name: frequency_penalty use_template: frequency_penalty pricing: diff --git a/api/core/model_runtime/model_providers/siliconflow/llm/gemma-2-9b-it.yaml b/api/core/model_runtime/model_providers/siliconflow/llm/gemma-2-9b-it.yaml index d7e19b46f6d..b096b9b647c 100644 --- a/api/core/model_runtime/model_providers/siliconflow/llm/gemma-2-9b-it.yaml +++ b/api/core/model_runtime/model_providers/siliconflow/llm/gemma-2-9b-it.yaml @@ -21,6 +21,15 @@ parameter_rules: en_US: Specifies the upper limit on the length of generated results. If the generated results are truncated, you can increase this parameter. - name: top_p use_template: top_p + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false - name: frequency_penalty use_template: frequency_penalty pricing: diff --git a/api/core/model_runtime/model_providers/siliconflow/llm/glm4-9b-chat.yaml b/api/core/model_runtime/model_providers/siliconflow/llm/glm4-9b-chat.yaml index 9b32a024774..87acc557b79 100644 --- a/api/core/model_runtime/model_providers/siliconflow/llm/glm4-9b-chat.yaml +++ b/api/core/model_runtime/model_providers/siliconflow/llm/glm4-9b-chat.yaml @@ -21,6 +21,15 @@ parameter_rules: en_US: Specifies the upper limit on the length of generated results. If the generated results are truncated, you can increase this parameter. - name: top_p use_template: top_p + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false - name: frequency_penalty use_template: frequency_penalty pricing: diff --git a/api/core/model_runtime/model_providers/siliconflow/llm/internlm2_5-20b-chat.yaml b/api/core/model_runtime/model_providers/siliconflow/llm/internlm2_5-20b-chat.yaml index d9663582e5c..60157c2b46a 100644 --- a/api/core/model_runtime/model_providers/siliconflow/llm/internlm2_5-20b-chat.yaml +++ b/api/core/model_runtime/model_providers/siliconflow/llm/internlm2_5-20b-chat.yaml @@ -21,6 +21,15 @@ parameter_rules: en_US: Specifies the upper limit on the length of generated results. If the generated results are truncated, you can increase this parameter. - name: top_p use_template: top_p + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false - name: frequency_penalty use_template: frequency_penalty pricing: diff --git a/api/core/model_runtime/model_providers/siliconflow/llm/internlm2_5-7b-chat.yaml b/api/core/model_runtime/model_providers/siliconflow/llm/internlm2_5-7b-chat.yaml index 73ad4480aa2..faf4af7ea3f 100644 --- a/api/core/model_runtime/model_providers/siliconflow/llm/internlm2_5-7b-chat.yaml +++ b/api/core/model_runtime/model_providers/siliconflow/llm/internlm2_5-7b-chat.yaml @@ -21,6 +21,15 @@ parameter_rules: en_US: Specifies the upper limit on the length of generated results. If the generated results are truncated, you can increase this parameter. - name: top_p use_template: top_p + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false - name: frequency_penalty use_template: frequency_penalty pricing: diff --git a/api/core/model_runtime/model_providers/siliconflow/llm/meta-mlama-3-70b-instruct.yaml b/api/core/model_runtime/model_providers/siliconflow/llm/meta-mlama-3-70b-instruct.yaml index 9993d781ac8..d01770cb010 100644 --- a/api/core/model_runtime/model_providers/siliconflow/llm/meta-mlama-3-70b-instruct.yaml +++ b/api/core/model_runtime/model_providers/siliconflow/llm/meta-mlama-3-70b-instruct.yaml @@ -21,6 +21,15 @@ parameter_rules: en_US: Specifies the upper limit on the length of generated results. If the generated results are truncated, you can increase this parameter. - name: top_p use_template: top_p + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false - name: frequency_penalty use_template: frequency_penalty pricing: diff --git a/api/core/model_runtime/model_providers/siliconflow/llm/meta-mlama-3-8b-instruct.yaml b/api/core/model_runtime/model_providers/siliconflow/llm/meta-mlama-3-8b-instruct.yaml index 60e3764789e..3cd75d89e8b 100644 --- a/api/core/model_runtime/model_providers/siliconflow/llm/meta-mlama-3-8b-instruct.yaml +++ b/api/core/model_runtime/model_providers/siliconflow/llm/meta-mlama-3-8b-instruct.yaml @@ -21,6 +21,15 @@ parameter_rules: en_US: Specifies the upper limit on the length of generated results. If the generated results are truncated, you can increase this parameter. - name: top_p use_template: top_p + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false - name: frequency_penalty use_template: frequency_penalty pricing: diff --git a/api/core/model_runtime/model_providers/siliconflow/llm/meta-mlama-3.1-405b-instruct.yaml b/api/core/model_runtime/model_providers/siliconflow/llm/meta-mlama-3.1-405b-instruct.yaml index f992660aa2e..3506a70bccf 100644 --- a/api/core/model_runtime/model_providers/siliconflow/llm/meta-mlama-3.1-405b-instruct.yaml +++ b/api/core/model_runtime/model_providers/siliconflow/llm/meta-mlama-3.1-405b-instruct.yaml @@ -21,6 +21,15 @@ parameter_rules: en_US: Specifies the upper limit on the length of generated results. If the generated results are truncated, you can increase this parameter. - name: top_p use_template: top_p + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false - name: frequency_penalty use_template: frequency_penalty pricing: diff --git a/api/core/model_runtime/model_providers/siliconflow/llm/meta-mlama-3.1-70b-instruct.yaml b/api/core/model_runtime/model_providers/siliconflow/llm/meta-mlama-3.1-70b-instruct.yaml index 1c69d63a400..994a754a82b 100644 --- a/api/core/model_runtime/model_providers/siliconflow/llm/meta-mlama-3.1-70b-instruct.yaml +++ b/api/core/model_runtime/model_providers/siliconflow/llm/meta-mlama-3.1-70b-instruct.yaml @@ -21,6 +21,15 @@ parameter_rules: en_US: Specifies the upper limit on the length of generated results. If the generated results are truncated, you can increase this parameter. - name: top_p use_template: top_p + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false - name: frequency_penalty use_template: frequency_penalty pricing: diff --git a/api/core/model_runtime/model_providers/siliconflow/llm/meta-mlama-3.1-8b-instruct.yaml b/api/core/model_runtime/model_providers/siliconflow/llm/meta-mlama-3.1-8b-instruct.yaml index a97002a5ca3..ebfa9aac9d9 100644 --- a/api/core/model_runtime/model_providers/siliconflow/llm/meta-mlama-3.1-8b-instruct.yaml +++ b/api/core/model_runtime/model_providers/siliconflow/llm/meta-mlama-3.1-8b-instruct.yaml @@ -21,6 +21,15 @@ parameter_rules: en_US: Specifies the upper limit on the length of generated results. If the generated results are truncated, you can increase this parameter. - name: top_p use_template: top_p + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false - name: frequency_penalty use_template: frequency_penalty pricing: diff --git a/api/core/model_runtime/model_providers/siliconflow/llm/mistral-7b-instruct-v0.2.yaml b/api/core/model_runtime/model_providers/siliconflow/llm/mistral-7b-instruct-v0.2.yaml index 89fb153ba09..a71d8688a87 100644 --- a/api/core/model_runtime/model_providers/siliconflow/llm/mistral-7b-instruct-v0.2.yaml +++ b/api/core/model_runtime/model_providers/siliconflow/llm/mistral-7b-instruct-v0.2.yaml @@ -21,6 +21,15 @@ parameter_rules: en_US: Specifies the upper limit on the length of generated results. If the generated results are truncated, you can increase this parameter. - name: top_p use_template: top_p + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false - name: frequency_penalty use_template: frequency_penalty pricing: diff --git a/api/core/model_runtime/model_providers/siliconflow/llm/mistral-8x7b-instruct-v0.1.yaml b/api/core/model_runtime/model_providers/siliconflow/llm/mistral-8x7b-instruct-v0.1.yaml index 2785e7496fb..db45a75c6de 100644 --- a/api/core/model_runtime/model_providers/siliconflow/llm/mistral-8x7b-instruct-v0.1.yaml +++ b/api/core/model_runtime/model_providers/siliconflow/llm/mistral-8x7b-instruct-v0.1.yaml @@ -21,6 +21,15 @@ parameter_rules: en_US: Specifies the upper limit on the length of generated results. If the generated results are truncated, you can increase this parameter. - name: top_p use_template: top_p + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false - name: frequency_penalty use_template: frequency_penalty pricing: diff --git a/api/core/model_runtime/model_providers/siliconflow/llm/qwen2-1.5b-instruct.yaml b/api/core/model_runtime/model_providers/siliconflow/llm/qwen2-1.5b-instruct.yaml index f6c976af8e7..bec5d37c57b 100644 --- a/api/core/model_runtime/model_providers/siliconflow/llm/qwen2-1.5b-instruct.yaml +++ b/api/core/model_runtime/model_providers/siliconflow/llm/qwen2-1.5b-instruct.yaml @@ -21,6 +21,15 @@ parameter_rules: en_US: Specifies the upper limit on the length of generated results. If the generated results are truncated, you can increase this parameter. - name: top_p use_template: top_p + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false - name: frequency_penalty use_template: frequency_penalty pricing: diff --git a/api/core/model_runtime/model_providers/siliconflow/llm/qwen2-57b-a14b-instruct.yaml b/api/core/model_runtime/model_providers/siliconflow/llm/qwen2-57b-a14b-instruct.yaml index a996e919ea9..b2461335f84 100644 --- a/api/core/model_runtime/model_providers/siliconflow/llm/qwen2-57b-a14b-instruct.yaml +++ b/api/core/model_runtime/model_providers/siliconflow/llm/qwen2-57b-a14b-instruct.yaml @@ -21,6 +21,15 @@ parameter_rules: en_US: Specifies the upper limit on the length of generated results. If the generated results are truncated, you can increase this parameter. - name: top_p use_template: top_p + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false - name: frequency_penalty use_template: frequency_penalty pricing: diff --git a/api/core/model_runtime/model_providers/siliconflow/llm/qwen2-72b-instruct.yaml b/api/core/model_runtime/model_providers/siliconflow/llm/qwen2-72b-instruct.yaml index a6e2c22dac8..e0f23bd89eb 100644 --- a/api/core/model_runtime/model_providers/siliconflow/llm/qwen2-72b-instruct.yaml +++ b/api/core/model_runtime/model_providers/siliconflow/llm/qwen2-72b-instruct.yaml @@ -21,6 +21,15 @@ parameter_rules: en_US: Specifies the upper limit on the length of generated results. If the generated results are truncated, you can increase this parameter. - name: top_p use_template: top_p + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false - name: frequency_penalty use_template: frequency_penalty pricing: diff --git a/api/core/model_runtime/model_providers/siliconflow/llm/qwen2-7b-instruct.yaml b/api/core/model_runtime/model_providers/siliconflow/llm/qwen2-7b-instruct.yaml index d8bea5e1292..47a9da8119c 100644 --- a/api/core/model_runtime/model_providers/siliconflow/llm/qwen2-7b-instruct.yaml +++ b/api/core/model_runtime/model_providers/siliconflow/llm/qwen2-7b-instruct.yaml @@ -21,6 +21,15 @@ parameter_rules: en_US: Specifies the upper limit on the length of generated results. If the generated results are truncated, you can increase this parameter. - name: top_p use_template: top_p + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false - name: frequency_penalty use_template: frequency_penalty pricing: diff --git a/api/core/model_runtime/model_providers/siliconflow/llm/qwen2.5-14b-instruct.yaml b/api/core/model_runtime/model_providers/siliconflow/llm/qwen2.5-14b-instruct.yaml index 02a401464bb..9cc5ac4c91d 100644 --- a/api/core/model_runtime/model_providers/siliconflow/llm/qwen2.5-14b-instruct.yaml +++ b/api/core/model_runtime/model_providers/siliconflow/llm/qwen2.5-14b-instruct.yaml @@ -21,6 +21,15 @@ parameter_rules: en_US: Specifies the upper limit on the length of generated results. If the generated results are truncated, you can increase this parameter. - name: top_p use_template: top_p + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false - name: frequency_penalty use_template: frequency_penalty pricing: diff --git a/api/core/model_runtime/model_providers/siliconflow/llm/qwen2.5-32b-instruct.yaml b/api/core/model_runtime/model_providers/siliconflow/llm/qwen2.5-32b-instruct.yaml index d084617e7d7..c7fb21e9e10 100644 --- a/api/core/model_runtime/model_providers/siliconflow/llm/qwen2.5-32b-instruct.yaml +++ b/api/core/model_runtime/model_providers/siliconflow/llm/qwen2.5-32b-instruct.yaml @@ -21,6 +21,15 @@ parameter_rules: en_US: Specifies the upper limit on the length of generated results. If the generated results are truncated, you can increase this parameter. - name: top_p use_template: top_p + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false - name: frequency_penalty use_template: frequency_penalty pricing: diff --git a/api/core/model_runtime/model_providers/siliconflow/llm/qwen2.5-72b-instruct.yaml b/api/core/model_runtime/model_providers/siliconflow/llm/qwen2.5-72b-instruct.yaml index dfbad2494cd..03136c88a11 100644 --- a/api/core/model_runtime/model_providers/siliconflow/llm/qwen2.5-72b-instruct.yaml +++ b/api/core/model_runtime/model_providers/siliconflow/llm/qwen2.5-72b-instruct.yaml @@ -21,6 +21,15 @@ parameter_rules: en_US: Specifies the upper limit on the length of generated results. If the generated results are truncated, you can increase this parameter. - name: top_p use_template: top_p + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false - name: frequency_penalty use_template: frequency_penalty pricing: diff --git a/api/core/model_runtime/model_providers/siliconflow/llm/qwen2.5-7b-instruct.yaml b/api/core/model_runtime/model_providers/siliconflow/llm/qwen2.5-7b-instruct.yaml index cdc8ffc4d21..99412adde7a 100644 --- a/api/core/model_runtime/model_providers/siliconflow/llm/qwen2.5-7b-instruct.yaml +++ b/api/core/model_runtime/model_providers/siliconflow/llm/qwen2.5-7b-instruct.yaml @@ -21,6 +21,15 @@ parameter_rules: en_US: Specifies the upper limit on the length of generated results. If the generated results are truncated, you can increase this parameter. - name: top_p use_template: top_p + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false - name: frequency_penalty use_template: frequency_penalty pricing: diff --git a/api/core/model_runtime/model_providers/siliconflow/llm/yi-1.5-34b-chat.yaml b/api/core/model_runtime/model_providers/siliconflow/llm/yi-1.5-34b-chat.yaml index 864ba46f1ad..3e25f82369f 100644 --- a/api/core/model_runtime/model_providers/siliconflow/llm/yi-1.5-34b-chat.yaml +++ b/api/core/model_runtime/model_providers/siliconflow/llm/yi-1.5-34b-chat.yaml @@ -21,6 +21,15 @@ parameter_rules: en_US: Specifies the upper limit on the length of generated results. If the generated results are truncated, you can increase this parameter. - name: top_p use_template: top_p + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false - name: frequency_penalty use_template: frequency_penalty pricing: diff --git a/api/core/model_runtime/model_providers/siliconflow/llm/yi-1.5-6b-chat.yaml b/api/core/model_runtime/model_providers/siliconflow/llm/yi-1.5-6b-chat.yaml index fe4c8b4b3e0..827b2ce1e5d 100644 --- a/api/core/model_runtime/model_providers/siliconflow/llm/yi-1.5-6b-chat.yaml +++ b/api/core/model_runtime/model_providers/siliconflow/llm/yi-1.5-6b-chat.yaml @@ -21,6 +21,15 @@ parameter_rules: en_US: Specifies the upper limit on the length of generated results. If the generated results are truncated, you can increase this parameter. - name: top_p use_template: top_p + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false - name: frequency_penalty use_template: frequency_penalty pricing: diff --git a/api/core/model_runtime/model_providers/siliconflow/llm/yi-1.5-9b-chat.yaml b/api/core/model_runtime/model_providers/siliconflow/llm/yi-1.5-9b-chat.yaml index c61f0dc53fe..112fcbfe97b 100644 --- a/api/core/model_runtime/model_providers/siliconflow/llm/yi-1.5-9b-chat.yaml +++ b/api/core/model_runtime/model_providers/siliconflow/llm/yi-1.5-9b-chat.yaml @@ -21,6 +21,15 @@ parameter_rules: en_US: Specifies the upper limit on the length of generated results. If the generated results are truncated, you can increase this parameter. - name: top_p use_template: top_p + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false - name: frequency_penalty use_template: frequency_penalty pricing: