modelparams 0.0.32__tar.gz → 0.0.34__tar.gz
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.
- {modelparams-0.0.32 → modelparams-0.0.34}/PKG-INFO +1 -1
- {modelparams-0.0.32 → modelparams-0.0.34}/pyproject.toml +1 -1
- {modelparams-0.0.32 → modelparams-0.0.34}/src/modelparams/_generated/catalog.json +482 -1
- {modelparams-0.0.32 → modelparams-0.0.34}/src/modelparams/_generated/model_ids.py +2 -0
- {modelparams-0.0.32 → modelparams-0.0.34}/src/modelparams/_generated/registry.py +1 -0
- {modelparams-0.0.32 → modelparams-0.0.34}/src/modelparams/types/alibaba.py +15 -0
- {modelparams-0.0.32 → modelparams-0.0.34}/src/modelparams/types/moonshot.py +37 -0
- {modelparams-0.0.32 → modelparams-0.0.34}/.gitignore +0 -0
- {modelparams-0.0.32 → modelparams-0.0.34}/LICENSE +0 -0
- {modelparams-0.0.32 → modelparams-0.0.34}/README.md +0 -0
- {modelparams-0.0.32 → modelparams-0.0.34}/src/modelparams/__init__.py +0 -0
- {modelparams-0.0.32 → modelparams-0.0.34}/src/modelparams/_generated/__init__.py +0 -0
- {modelparams-0.0.32 → modelparams-0.0.34}/src/modelparams/catalog.py +0 -0
- {modelparams-0.0.32 → modelparams-0.0.34}/src/modelparams/models.py +0 -0
- {modelparams-0.0.32 → modelparams-0.0.34}/src/modelparams/py.typed +0 -0
- {modelparams-0.0.32 → modelparams-0.0.34}/src/modelparams/types/__init__.py +0 -0
- {modelparams-0.0.32 → modelparams-0.0.34}/src/modelparams/types/anthropic.py +0 -0
- {modelparams-0.0.32 → modelparams-0.0.34}/src/modelparams/types/cerebras.py +0 -0
- {modelparams-0.0.32 → modelparams-0.0.34}/src/modelparams/types/cohere.py +0 -0
- {modelparams-0.0.32 → modelparams-0.0.34}/src/modelparams/types/deepseek.py +0 -0
- {modelparams-0.0.32 → modelparams-0.0.34}/src/modelparams/types/google.py +0 -0
- {modelparams-0.0.32 → modelparams-0.0.34}/src/modelparams/types/groq.py +0 -0
- {modelparams-0.0.32 → modelparams-0.0.34}/src/modelparams/types/meta.py +0 -0
- {modelparams-0.0.32 → modelparams-0.0.34}/src/modelparams/types/minimax.py +0 -0
- {modelparams-0.0.32 → modelparams-0.0.34}/src/modelparams/types/mistral.py +0 -0
- {modelparams-0.0.32 → modelparams-0.0.34}/src/modelparams/types/nvidia.py +0 -0
- {modelparams-0.0.32 → modelparams-0.0.34}/src/modelparams/types/openai.py +0 -0
- {modelparams-0.0.32 → modelparams-0.0.34}/src/modelparams/types/perplexity.py +0 -0
- {modelparams-0.0.32 → modelparams-0.0.34}/src/modelparams/types/thinking_machines.py +0 -0
- {modelparams-0.0.32 → modelparams-0.0.34}/src/modelparams/types/xai.py +0 -0
- {modelparams-0.0.32 → modelparams-0.0.34}/src/modelparams/types/xiaomi.py +0 -0
- {modelparams-0.0.32 → modelparams-0.0.34}/src/modelparams/types/z_ai.py +0 -0
- {modelparams-0.0.32 → modelparams-0.0.34}/src/modelparams/validation.py +0 -0
|
@@ -1,6 +1,6 @@
|
|
|
1
1
|
Metadata-Version: 2.5
|
|
2
2
|
Name: modelparams
|
|
3
|
-
Version: 0.0.
|
|
3
|
+
Version: 0.0.34
|
|
4
4
|
Summary: Typed model parameters for Python, generated from the modelparams.dev catalog.
|
|
5
5
|
Project-URL: Homepage, https://modelparams.dev
|
|
6
6
|
Project-URL: Repository, https://github.com/mnfst/modelparameters.dev
|
|
@@ -449,6 +449,81 @@
|
|
|
449
449
|
}
|
|
450
450
|
]
|
|
451
451
|
},
|
|
452
|
+
{
|
|
453
|
+
"provider": "alibaba",
|
|
454
|
+
"authType": "api_key",
|
|
455
|
+
"model": "glm-5.2",
|
|
456
|
+
"params": [
|
|
457
|
+
{
|
|
458
|
+
"path": "max_completion_tokens",
|
|
459
|
+
"label": "Max tokens",
|
|
460
|
+
"description": "Maximum number of tokens to generate, including both reasoning and the final answer.",
|
|
461
|
+
"group": "generation_length",
|
|
462
|
+
"type": "integer",
|
|
463
|
+
"range": {
|
|
464
|
+
"min": 1
|
|
465
|
+
}
|
|
466
|
+
},
|
|
467
|
+
{
|
|
468
|
+
"path": "temperature",
|
|
469
|
+
"label": "Temperature",
|
|
470
|
+
"description": "Controls randomness. Lower values make outputs more focused; higher values make them more varied.",
|
|
471
|
+
"group": "sampling",
|
|
472
|
+
"type": "number",
|
|
473
|
+
"range": {
|
|
474
|
+
"min": 0,
|
|
475
|
+
"max": 1.9,
|
|
476
|
+
"step": 0.1
|
|
477
|
+
}
|
|
478
|
+
},
|
|
479
|
+
{
|
|
480
|
+
"path": "top_p",
|
|
481
|
+
"label": "Top P",
|
|
482
|
+
"description": "Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.",
|
|
483
|
+
"group": "sampling",
|
|
484
|
+
"type": "number",
|
|
485
|
+
"range": {
|
|
486
|
+
"min": 0,
|
|
487
|
+
"max": 1,
|
|
488
|
+
"step": 0.01
|
|
489
|
+
}
|
|
490
|
+
},
|
|
491
|
+
{
|
|
492
|
+
"path": "extra_body.top_k",
|
|
493
|
+
"label": "Top K",
|
|
494
|
+
"description": "Limits generation to the selected number of highest-probability tokens. Values above 100 disable top-k sampling.",
|
|
495
|
+
"group": "sampling",
|
|
496
|
+
"type": "integer",
|
|
497
|
+
"default": 20,
|
|
498
|
+
"range": {
|
|
499
|
+
"min": 0
|
|
500
|
+
}
|
|
501
|
+
},
|
|
502
|
+
{
|
|
503
|
+
"path": "extra_body.enable_thinking",
|
|
504
|
+
"label": "Enable thinking",
|
|
505
|
+
"description": "Toggles the model's hybrid thinking mode, sent as a provider-specific extra body field on OpenAI-compatible clients.",
|
|
506
|
+
"group": "reasoning",
|
|
507
|
+
"type": "boolean",
|
|
508
|
+
"default": true
|
|
509
|
+
},
|
|
510
|
+
{
|
|
511
|
+
"path": "extra_body.thinking_budget",
|
|
512
|
+
"label": "Thinking budget",
|
|
513
|
+
"description": "Maximum number of tokens the model may spend on reasoning before it starts the final answer; defaults to the model's maximum reasoning length.",
|
|
514
|
+
"group": "reasoning",
|
|
515
|
+
"applicability": {
|
|
516
|
+
"only": {
|
|
517
|
+
"extra_body.enable_thinking": true
|
|
518
|
+
}
|
|
519
|
+
},
|
|
520
|
+
"type": "integer",
|
|
521
|
+
"range": {
|
|
522
|
+
"min": 1
|
|
523
|
+
}
|
|
524
|
+
}
|
|
525
|
+
]
|
|
526
|
+
},
|
|
452
527
|
{
|
|
453
528
|
"provider": "alibaba",
|
|
454
529
|
"authType": "api_key",
|
|
@@ -14648,12 +14723,20 @@
|
|
|
14648
14723
|
"min": 1
|
|
14649
14724
|
}
|
|
14650
14725
|
},
|
|
14726
|
+
{
|
|
14727
|
+
"path": "stop",
|
|
14728
|
+
"label": "Stop sequence",
|
|
14729
|
+
"description": "Stops generation when this sequence is produced. Moonshot accepts up to 5 sequences of at most 32 bytes each.",
|
|
14730
|
+
"group": "generation_length",
|
|
14731
|
+
"type": "string"
|
|
14732
|
+
},
|
|
14651
14733
|
{
|
|
14652
14734
|
"path": "thinking.type",
|
|
14653
14735
|
"label": "Thinking mode",
|
|
14654
|
-
"description": "Controls whether Kimi reasons step by step before answering
|
|
14736
|
+
"description": "Controls whether Kimi reasons step by step before answering. Thinking is enabled by default; set disabled to respond directly.",
|
|
14655
14737
|
"group": "reasoning",
|
|
14656
14738
|
"type": "enum",
|
|
14739
|
+
"default": "enabled",
|
|
14657
14740
|
"values": [
|
|
14658
14741
|
"enabled",
|
|
14659
14742
|
"disabled"
|
|
@@ -14670,6 +14753,42 @@
|
|
|
14670
14753
|
"text",
|
|
14671
14754
|
"json_object"
|
|
14672
14755
|
]
|
|
14756
|
+
},
|
|
14757
|
+
{
|
|
14758
|
+
"path": "tool_choice",
|
|
14759
|
+
"label": "Tool choice",
|
|
14760
|
+
"description": "Controls tool calling: auto lets the model decide and none blocks tool calls. Moonshot rejects required while thinking is enabled, which is the default on this model.",
|
|
14761
|
+
"group": "tooling",
|
|
14762
|
+
"type": "enum",
|
|
14763
|
+
"default": "auto",
|
|
14764
|
+
"values": [
|
|
14765
|
+
"auto",
|
|
14766
|
+
"none"
|
|
14767
|
+
]
|
|
14768
|
+
},
|
|
14769
|
+
{
|
|
14770
|
+
"path": "logprobs",
|
|
14771
|
+
"label": "Log probabilities",
|
|
14772
|
+
"description": "Controls whether the response includes log probabilities for the generated tokens.",
|
|
14773
|
+
"group": "observability",
|
|
14774
|
+
"type": "boolean",
|
|
14775
|
+
"default": false
|
|
14776
|
+
},
|
|
14777
|
+
{
|
|
14778
|
+
"path": "top_logprobs",
|
|
14779
|
+
"label": "Top log probabilities",
|
|
14780
|
+
"description": "Number of most likely tokens to return a log probability for at each position.",
|
|
14781
|
+
"group": "observability",
|
|
14782
|
+
"applicability": {
|
|
14783
|
+
"only": {
|
|
14784
|
+
"logprobs": true
|
|
14785
|
+
}
|
|
14786
|
+
},
|
|
14787
|
+
"type": "integer",
|
|
14788
|
+
"range": {
|
|
14789
|
+
"min": 0,
|
|
14790
|
+
"max": 20
|
|
14791
|
+
}
|
|
14673
14792
|
}
|
|
14674
14793
|
]
|
|
14675
14794
|
},
|
|
@@ -14688,6 +14807,13 @@
|
|
|
14688
14807
|
"min": 1
|
|
14689
14808
|
}
|
|
14690
14809
|
},
|
|
14810
|
+
{
|
|
14811
|
+
"path": "stop",
|
|
14812
|
+
"label": "Stop sequence",
|
|
14813
|
+
"description": "Stops generation when this sequence is produced. Moonshot accepts up to 5 sequences of at most 32 bytes each.",
|
|
14814
|
+
"group": "generation_length",
|
|
14815
|
+
"type": "string"
|
|
14816
|
+
},
|
|
14691
14817
|
{
|
|
14692
14818
|
"path": "thinking.type",
|
|
14693
14819
|
"label": "Thinking mode",
|
|
@@ -14700,6 +14826,18 @@
|
|
|
14700
14826
|
"disabled"
|
|
14701
14827
|
]
|
|
14702
14828
|
},
|
|
14829
|
+
{
|
|
14830
|
+
"path": "thinking.keep",
|
|
14831
|
+
"label": "Keep prior reasoning",
|
|
14832
|
+
"description": "Set to all to carry reasoning content from earlier assistant turns into the request; leave null to send only the final answers.",
|
|
14833
|
+
"group": "reasoning",
|
|
14834
|
+
"type": "enum",
|
|
14835
|
+
"default": null,
|
|
14836
|
+
"values": [
|
|
14837
|
+
"all",
|
|
14838
|
+
null
|
|
14839
|
+
]
|
|
14840
|
+
},
|
|
14703
14841
|
{
|
|
14704
14842
|
"path": "response_format.type",
|
|
14705
14843
|
"label": "Response format",
|
|
@@ -14711,6 +14849,42 @@
|
|
|
14711
14849
|
"text",
|
|
14712
14850
|
"json_object"
|
|
14713
14851
|
]
|
|
14852
|
+
},
|
|
14853
|
+
{
|
|
14854
|
+
"path": "tool_choice",
|
|
14855
|
+
"label": "Tool choice",
|
|
14856
|
+
"description": "Controls tool calling: auto lets the model decide and none blocks tool calls. Moonshot rejects required while thinking is enabled, which is the default on this model.",
|
|
14857
|
+
"group": "tooling",
|
|
14858
|
+
"type": "enum",
|
|
14859
|
+
"default": "auto",
|
|
14860
|
+
"values": [
|
|
14861
|
+
"auto",
|
|
14862
|
+
"none"
|
|
14863
|
+
]
|
|
14864
|
+
},
|
|
14865
|
+
{
|
|
14866
|
+
"path": "logprobs",
|
|
14867
|
+
"label": "Log probabilities",
|
|
14868
|
+
"description": "Controls whether the response includes log probabilities for the generated tokens.",
|
|
14869
|
+
"group": "observability",
|
|
14870
|
+
"type": "boolean",
|
|
14871
|
+
"default": false
|
|
14872
|
+
},
|
|
14873
|
+
{
|
|
14874
|
+
"path": "top_logprobs",
|
|
14875
|
+
"label": "Top log probabilities",
|
|
14876
|
+
"description": "Number of most likely tokens to return a log probability for at each position.",
|
|
14877
|
+
"group": "observability",
|
|
14878
|
+
"applicability": {
|
|
14879
|
+
"only": {
|
|
14880
|
+
"logprobs": true
|
|
14881
|
+
}
|
|
14882
|
+
},
|
|
14883
|
+
"type": "integer",
|
|
14884
|
+
"range": {
|
|
14885
|
+
"min": 0,
|
|
14886
|
+
"max": 20
|
|
14887
|
+
}
|
|
14714
14888
|
}
|
|
14715
14889
|
]
|
|
14716
14890
|
},
|
|
@@ -14741,6 +14915,18 @@
|
|
|
14741
14915
|
"disabled"
|
|
14742
14916
|
]
|
|
14743
14917
|
},
|
|
14918
|
+
{
|
|
14919
|
+
"path": "thinking.keep",
|
|
14920
|
+
"label": "Keep prior reasoning",
|
|
14921
|
+
"description": "Set to all to carry reasoning content from earlier assistant turns into the request; leave null to send only the final answers.",
|
|
14922
|
+
"group": "reasoning",
|
|
14923
|
+
"type": "enum",
|
|
14924
|
+
"default": null,
|
|
14925
|
+
"values": [
|
|
14926
|
+
"all",
|
|
14927
|
+
null
|
|
14928
|
+
]
|
|
14929
|
+
},
|
|
14744
14930
|
{
|
|
14745
14931
|
"path": "response_format.type",
|
|
14746
14932
|
"label": "Response format",
|
|
@@ -14771,6 +14957,24 @@
|
|
|
14771
14957
|
"min": 1
|
|
14772
14958
|
}
|
|
14773
14959
|
},
|
|
14960
|
+
{
|
|
14961
|
+
"path": "stop",
|
|
14962
|
+
"label": "Stop sequence",
|
|
14963
|
+
"description": "Stops generation when this sequence is produced. Moonshot accepts up to 5 sequences of at most 32 bytes each.",
|
|
14964
|
+
"group": "generation_length",
|
|
14965
|
+
"type": "string"
|
|
14966
|
+
},
|
|
14967
|
+
{
|
|
14968
|
+
"path": "thinking.type",
|
|
14969
|
+
"label": "Thinking mode",
|
|
14970
|
+
"description": "Thinking is always on for this model. Enabled is the only accepted value; disabled returns an error.",
|
|
14971
|
+
"group": "reasoning",
|
|
14972
|
+
"type": "enum",
|
|
14973
|
+
"default": "enabled",
|
|
14974
|
+
"values": [
|
|
14975
|
+
"enabled"
|
|
14976
|
+
]
|
|
14977
|
+
},
|
|
14774
14978
|
{
|
|
14775
14979
|
"path": "response_format.type",
|
|
14776
14980
|
"label": "Response format",
|
|
@@ -14783,6 +14987,42 @@
|
|
|
14783
14987
|
"json_object",
|
|
14784
14988
|
"json_schema"
|
|
14785
14989
|
]
|
|
14990
|
+
},
|
|
14991
|
+
{
|
|
14992
|
+
"path": "tool_choice",
|
|
14993
|
+
"label": "Tool choice",
|
|
14994
|
+
"description": "Controls tool calling: auto lets the model decide and none blocks tool calls. Moonshot rejects required while thinking is enabled, which is the default on this model.",
|
|
14995
|
+
"group": "tooling",
|
|
14996
|
+
"type": "enum",
|
|
14997
|
+
"default": "auto",
|
|
14998
|
+
"values": [
|
|
14999
|
+
"auto",
|
|
15000
|
+
"none"
|
|
15001
|
+
]
|
|
15002
|
+
},
|
|
15003
|
+
{
|
|
15004
|
+
"path": "logprobs",
|
|
15005
|
+
"label": "Log probabilities",
|
|
15006
|
+
"description": "Controls whether the response includes log probabilities for the generated tokens.",
|
|
15007
|
+
"group": "observability",
|
|
15008
|
+
"type": "boolean",
|
|
15009
|
+
"default": false
|
|
15010
|
+
},
|
|
15011
|
+
{
|
|
15012
|
+
"path": "top_logprobs",
|
|
15013
|
+
"label": "Top log probabilities",
|
|
15014
|
+
"description": "Number of most likely tokens to return a log probability for at each position.",
|
|
15015
|
+
"group": "observability",
|
|
15016
|
+
"applicability": {
|
|
15017
|
+
"only": {
|
|
15018
|
+
"logprobs": true
|
|
15019
|
+
}
|
|
15020
|
+
},
|
|
15021
|
+
"type": "integer",
|
|
15022
|
+
"range": {
|
|
15023
|
+
"min": 0,
|
|
15024
|
+
"max": 20
|
|
15025
|
+
}
|
|
14786
15026
|
}
|
|
14787
15027
|
]
|
|
14788
15028
|
},
|
|
@@ -14802,6 +15042,24 @@
|
|
|
14802
15042
|
"min": 1
|
|
14803
15043
|
}
|
|
14804
15044
|
},
|
|
15045
|
+
{
|
|
15046
|
+
"path": "stop",
|
|
15047
|
+
"label": "Stop sequence",
|
|
15048
|
+
"description": "Stops generation when this sequence is produced. Moonshot accepts up to 5 sequences of at most 32 bytes each.",
|
|
15049
|
+
"group": "generation_length",
|
|
15050
|
+
"type": "string"
|
|
15051
|
+
},
|
|
15052
|
+
{
|
|
15053
|
+
"path": "thinking.type",
|
|
15054
|
+
"label": "Thinking mode",
|
|
15055
|
+
"description": "Thinking is always on for this model. Enabled is the only accepted value; disabled returns an error.",
|
|
15056
|
+
"group": "reasoning",
|
|
15057
|
+
"type": "enum",
|
|
15058
|
+
"default": "enabled",
|
|
15059
|
+
"values": [
|
|
15060
|
+
"enabled"
|
|
15061
|
+
]
|
|
15062
|
+
},
|
|
14805
15063
|
{
|
|
14806
15064
|
"path": "response_format.type",
|
|
14807
15065
|
"label": "Response format",
|
|
@@ -14814,6 +15072,42 @@
|
|
|
14814
15072
|
"json_object",
|
|
14815
15073
|
"json_schema"
|
|
14816
15074
|
]
|
|
15075
|
+
},
|
|
15076
|
+
{
|
|
15077
|
+
"path": "tool_choice",
|
|
15078
|
+
"label": "Tool choice",
|
|
15079
|
+
"description": "Controls tool calling: auto lets the model decide and none blocks tool calls. Moonshot rejects required while thinking is enabled, which is the default on this model.",
|
|
15080
|
+
"group": "tooling",
|
|
15081
|
+
"type": "enum",
|
|
15082
|
+
"default": "auto",
|
|
15083
|
+
"values": [
|
|
15084
|
+
"auto",
|
|
15085
|
+
"none"
|
|
15086
|
+
]
|
|
15087
|
+
},
|
|
15088
|
+
{
|
|
15089
|
+
"path": "logprobs",
|
|
15090
|
+
"label": "Log probabilities",
|
|
15091
|
+
"description": "Controls whether the response includes log probabilities for the generated tokens.",
|
|
15092
|
+
"group": "observability",
|
|
15093
|
+
"type": "boolean",
|
|
15094
|
+
"default": false
|
|
15095
|
+
},
|
|
15096
|
+
{
|
|
15097
|
+
"path": "top_logprobs",
|
|
15098
|
+
"label": "Top log probabilities",
|
|
15099
|
+
"description": "Number of most likely tokens to return a log probability for at each position.",
|
|
15100
|
+
"group": "observability",
|
|
15101
|
+
"applicability": {
|
|
15102
|
+
"only": {
|
|
15103
|
+
"logprobs": true
|
|
15104
|
+
}
|
|
15105
|
+
},
|
|
15106
|
+
"type": "integer",
|
|
15107
|
+
"range": {
|
|
15108
|
+
"min": 0,
|
|
15109
|
+
"max": 20
|
|
15110
|
+
}
|
|
14817
15111
|
}
|
|
14818
15112
|
]
|
|
14819
15113
|
},
|
|
@@ -14832,6 +15126,17 @@
|
|
|
14832
15126
|
"min": 1
|
|
14833
15127
|
}
|
|
14834
15128
|
},
|
|
15129
|
+
{
|
|
15130
|
+
"path": "thinking.type",
|
|
15131
|
+
"label": "Thinking mode",
|
|
15132
|
+
"description": "Thinking is always on for this model. Enabled is the only accepted value; disabled returns an error.",
|
|
15133
|
+
"group": "reasoning",
|
|
15134
|
+
"type": "enum",
|
|
15135
|
+
"default": "enabled",
|
|
15136
|
+
"values": [
|
|
15137
|
+
"enabled"
|
|
15138
|
+
]
|
|
15139
|
+
},
|
|
14835
15140
|
{
|
|
14836
15141
|
"path": "response_format.type",
|
|
14837
15142
|
"label": "Response format",
|
|
@@ -14861,6 +15166,17 @@
|
|
|
14861
15166
|
"min": 1
|
|
14862
15167
|
}
|
|
14863
15168
|
},
|
|
15169
|
+
{
|
|
15170
|
+
"path": "thinking.type",
|
|
15171
|
+
"label": "Thinking mode",
|
|
15172
|
+
"description": "Thinking is always on for this model. Enabled is the only accepted value; disabled returns an error.",
|
|
15173
|
+
"group": "reasoning",
|
|
15174
|
+
"type": "enum",
|
|
15175
|
+
"default": "enabled",
|
|
15176
|
+
"values": [
|
|
15177
|
+
"enabled"
|
|
15178
|
+
]
|
|
15179
|
+
},
|
|
14864
15180
|
{
|
|
14865
15181
|
"path": "response_format.type",
|
|
14866
15182
|
"label": "Response format",
|
|
@@ -14890,6 +15206,13 @@
|
|
|
14890
15206
|
"min": 1
|
|
14891
15207
|
}
|
|
14892
15208
|
},
|
|
15209
|
+
{
|
|
15210
|
+
"path": "stop",
|
|
15211
|
+
"label": "Stop sequence",
|
|
15212
|
+
"description": "Stops generation when this sequence is produced. Moonshot accepts up to 5 sequences of at most 32 bytes each.",
|
|
15213
|
+
"group": "generation_length",
|
|
15214
|
+
"type": "string"
|
|
15215
|
+
},
|
|
14893
15216
|
{
|
|
14894
15217
|
"path": "thinking.type",
|
|
14895
15218
|
"label": "Thinking mode",
|
|
@@ -14901,6 +15224,19 @@
|
|
|
14901
15224
|
"disabled"
|
|
14902
15225
|
]
|
|
14903
15226
|
},
|
|
15227
|
+
{
|
|
15228
|
+
"path": "reasoning_effort",
|
|
15229
|
+
"label": "Reasoning effort",
|
|
15230
|
+
"description": "Controls how much reasoning Kimi performs before answering. Thinking is always on for this model, so it cannot be turned off.",
|
|
15231
|
+
"group": "reasoning",
|
|
15232
|
+
"type": "enum",
|
|
15233
|
+
"default": "max",
|
|
15234
|
+
"values": [
|
|
15235
|
+
"low",
|
|
15236
|
+
"high",
|
|
15237
|
+
"max"
|
|
15238
|
+
]
|
|
15239
|
+
},
|
|
14904
15240
|
{
|
|
14905
15241
|
"path": "response_format.type",
|
|
14906
15242
|
"label": "Response format",
|
|
@@ -14913,6 +15249,19 @@
|
|
|
14913
15249
|
"json_object",
|
|
14914
15250
|
"json_schema"
|
|
14915
15251
|
]
|
|
15252
|
+
},
|
|
15253
|
+
{
|
|
15254
|
+
"path": "tool_choice",
|
|
15255
|
+
"label": "Tool choice",
|
|
15256
|
+
"description": "Controls tool calling: auto lets the model decide, none blocks tool calls, and required forces one.",
|
|
15257
|
+
"group": "tooling",
|
|
15258
|
+
"type": "enum",
|
|
15259
|
+
"default": "auto",
|
|
15260
|
+
"values": [
|
|
15261
|
+
"auto",
|
|
15262
|
+
"none",
|
|
15263
|
+
"required"
|
|
15264
|
+
]
|
|
14916
15265
|
}
|
|
14917
15266
|
]
|
|
14918
15267
|
},
|
|
@@ -14931,6 +15280,13 @@
|
|
|
14931
15280
|
"min": 1
|
|
14932
15281
|
}
|
|
14933
15282
|
},
|
|
15283
|
+
{
|
|
15284
|
+
"path": "stop",
|
|
15285
|
+
"label": "Stop sequence",
|
|
15286
|
+
"description": "Stops generation when this sequence is produced. Moonshot accepts up to 5 sequences of at most 32 bytes each.",
|
|
15287
|
+
"group": "generation_length",
|
|
15288
|
+
"type": "string"
|
|
15289
|
+
},
|
|
14934
15290
|
{
|
|
14935
15291
|
"path": "temperature",
|
|
14936
15292
|
"label": "Temperature",
|
|
@@ -15006,6 +15362,43 @@
|
|
|
15006
15362
|
"text",
|
|
15007
15363
|
"json_object"
|
|
15008
15364
|
]
|
|
15365
|
+
},
|
|
15366
|
+
{
|
|
15367
|
+
"path": "tool_choice",
|
|
15368
|
+
"label": "Tool choice",
|
|
15369
|
+
"description": "Controls tool calling: auto lets the model decide, none blocks tool calls, and required forces one.",
|
|
15370
|
+
"group": "tooling",
|
|
15371
|
+
"type": "enum",
|
|
15372
|
+
"default": "auto",
|
|
15373
|
+
"values": [
|
|
15374
|
+
"auto",
|
|
15375
|
+
"none",
|
|
15376
|
+
"required"
|
|
15377
|
+
]
|
|
15378
|
+
},
|
|
15379
|
+
{
|
|
15380
|
+
"path": "logprobs",
|
|
15381
|
+
"label": "Log probabilities",
|
|
15382
|
+
"description": "Controls whether the response includes log probabilities for the generated tokens.",
|
|
15383
|
+
"group": "observability",
|
|
15384
|
+
"type": "boolean",
|
|
15385
|
+
"default": false
|
|
15386
|
+
},
|
|
15387
|
+
{
|
|
15388
|
+
"path": "top_logprobs",
|
|
15389
|
+
"label": "Top log probabilities",
|
|
15390
|
+
"description": "Number of most likely tokens to return a log probability for at each position.",
|
|
15391
|
+
"group": "observability",
|
|
15392
|
+
"applicability": {
|
|
15393
|
+
"only": {
|
|
15394
|
+
"logprobs": true
|
|
15395
|
+
}
|
|
15396
|
+
},
|
|
15397
|
+
"type": "integer",
|
|
15398
|
+
"range": {
|
|
15399
|
+
"min": 0,
|
|
15400
|
+
"max": 20
|
|
15401
|
+
}
|
|
15009
15402
|
}
|
|
15010
15403
|
]
|
|
15011
15404
|
},
|
|
@@ -15024,6 +15417,13 @@
|
|
|
15024
15417
|
"min": 1
|
|
15025
15418
|
}
|
|
15026
15419
|
},
|
|
15420
|
+
{
|
|
15421
|
+
"path": "stop",
|
|
15422
|
+
"label": "Stop sequence",
|
|
15423
|
+
"description": "Stops generation when this sequence is produced. Moonshot accepts up to 5 sequences of at most 32 bytes each.",
|
|
15424
|
+
"group": "generation_length",
|
|
15425
|
+
"type": "string"
|
|
15426
|
+
},
|
|
15027
15427
|
{
|
|
15028
15428
|
"path": "temperature",
|
|
15029
15429
|
"label": "Temperature",
|
|
@@ -15099,6 +15499,43 @@
|
|
|
15099
15499
|
"text",
|
|
15100
15500
|
"json_object"
|
|
15101
15501
|
]
|
|
15502
|
+
},
|
|
15503
|
+
{
|
|
15504
|
+
"path": "tool_choice",
|
|
15505
|
+
"label": "Tool choice",
|
|
15506
|
+
"description": "Controls tool calling: auto lets the model decide, none blocks tool calls, and required forces one.",
|
|
15507
|
+
"group": "tooling",
|
|
15508
|
+
"type": "enum",
|
|
15509
|
+
"default": "auto",
|
|
15510
|
+
"values": [
|
|
15511
|
+
"auto",
|
|
15512
|
+
"none",
|
|
15513
|
+
"required"
|
|
15514
|
+
]
|
|
15515
|
+
},
|
|
15516
|
+
{
|
|
15517
|
+
"path": "logprobs",
|
|
15518
|
+
"label": "Log probabilities",
|
|
15519
|
+
"description": "Controls whether the response includes log probabilities for the generated tokens.",
|
|
15520
|
+
"group": "observability",
|
|
15521
|
+
"type": "boolean",
|
|
15522
|
+
"default": false
|
|
15523
|
+
},
|
|
15524
|
+
{
|
|
15525
|
+
"path": "top_logprobs",
|
|
15526
|
+
"label": "Top log probabilities",
|
|
15527
|
+
"description": "Number of most likely tokens to return a log probability for at each position.",
|
|
15528
|
+
"group": "observability",
|
|
15529
|
+
"applicability": {
|
|
15530
|
+
"only": {
|
|
15531
|
+
"logprobs": true
|
|
15532
|
+
}
|
|
15533
|
+
},
|
|
15534
|
+
"type": "integer",
|
|
15535
|
+
"range": {
|
|
15536
|
+
"min": 0,
|
|
15537
|
+
"max": 20
|
|
15538
|
+
}
|
|
15102
15539
|
}
|
|
15103
15540
|
]
|
|
15104
15541
|
},
|
|
@@ -15117,6 +15554,13 @@
|
|
|
15117
15554
|
"min": 1
|
|
15118
15555
|
}
|
|
15119
15556
|
},
|
|
15557
|
+
{
|
|
15558
|
+
"path": "stop",
|
|
15559
|
+
"label": "Stop sequence",
|
|
15560
|
+
"description": "Stops generation when this sequence is produced. Moonshot accepts up to 5 sequences of at most 32 bytes each.",
|
|
15561
|
+
"group": "generation_length",
|
|
15562
|
+
"type": "string"
|
|
15563
|
+
},
|
|
15120
15564
|
{
|
|
15121
15565
|
"path": "temperature",
|
|
15122
15566
|
"label": "Temperature",
|
|
@@ -15192,6 +15636,43 @@
|
|
|
15192
15636
|
"text",
|
|
15193
15637
|
"json_object"
|
|
15194
15638
|
]
|
|
15639
|
+
},
|
|
15640
|
+
{
|
|
15641
|
+
"path": "tool_choice",
|
|
15642
|
+
"label": "Tool choice",
|
|
15643
|
+
"description": "Controls tool calling: auto lets the model decide, none blocks tool calls, and required forces one.",
|
|
15644
|
+
"group": "tooling",
|
|
15645
|
+
"type": "enum",
|
|
15646
|
+
"default": "auto",
|
|
15647
|
+
"values": [
|
|
15648
|
+
"auto",
|
|
15649
|
+
"none",
|
|
15650
|
+
"required"
|
|
15651
|
+
]
|
|
15652
|
+
},
|
|
15653
|
+
{
|
|
15654
|
+
"path": "logprobs",
|
|
15655
|
+
"label": "Log probabilities",
|
|
15656
|
+
"description": "Controls whether the response includes log probabilities for the generated tokens.",
|
|
15657
|
+
"group": "observability",
|
|
15658
|
+
"type": "boolean",
|
|
15659
|
+
"default": false
|
|
15660
|
+
},
|
|
15661
|
+
{
|
|
15662
|
+
"path": "top_logprobs",
|
|
15663
|
+
"label": "Top log probabilities",
|
|
15664
|
+
"description": "Number of most likely tokens to return a log probability for at each position.",
|
|
15665
|
+
"group": "observability",
|
|
15666
|
+
"applicability": {
|
|
15667
|
+
"only": {
|
|
15668
|
+
"logprobs": true
|
|
15669
|
+
}
|
|
15670
|
+
},
|
|
15671
|
+
"type": "integer",
|
|
15672
|
+
"range": {
|
|
15673
|
+
"min": 0,
|
|
15674
|
+
"max": 20
|
|
15675
|
+
}
|
|
15195
15676
|
}
|
|
15196
15677
|
]
|
|
15197
15678
|
},
|
|
@@ -10,6 +10,7 @@ ModelId = Literal[
|
|
|
10
10
|
"alibaba/deepseek-v4-pro",
|
|
11
11
|
"alibaba/deepseek-v4-pro-0813",
|
|
12
12
|
"alibaba/glm-5.1",
|
|
13
|
+
"alibaba/glm-5.2",
|
|
13
14
|
"alibaba/kimi-k2.7-code",
|
|
14
15
|
"alibaba/qwen-flash",
|
|
15
16
|
"alibaba/qwen-max",
|
|
@@ -301,6 +302,7 @@ MODEL_IDS: tuple[ModelId, ...] = (
|
|
|
301
302
|
"alibaba/deepseek-v4-pro",
|
|
302
303
|
"alibaba/deepseek-v4-pro-0813",
|
|
303
304
|
"alibaba/glm-5.1",
|
|
305
|
+
"alibaba/glm-5.2",
|
|
304
306
|
"alibaba/kimi-k2.7-code",
|
|
305
307
|
"alibaba/qwen-flash",
|
|
306
308
|
"alibaba/qwen-max",
|
|
@@ -31,6 +31,7 @@ PARAM_TYPES: dict[ModelId, Any] = {
|
|
|
31
31
|
"alibaba/deepseek-v4-pro": alibaba.Deepseek_V4_ProParams,
|
|
32
32
|
"alibaba/deepseek-v4-pro-0813": alibaba.Deepseek_V4_Pro_0813Params,
|
|
33
33
|
"alibaba/glm-5.1": alibaba.Glm_5_1Params,
|
|
34
|
+
"alibaba/glm-5.2": alibaba.Glm_5_2Params,
|
|
34
35
|
"alibaba/kimi-k2.7-code": alibaba.Kimi_K2_7_CodeParams,
|
|
35
36
|
"alibaba/qwen-flash": alibaba.Qwen_FlashParams,
|
|
36
37
|
"alibaba/qwen-max": alibaba.Qwen_MaxParams,
|
|
@@ -94,6 +94,20 @@ Glm_5_1Params = TypedDict(
|
|
|
94
94
|
)
|
|
95
95
|
setattr(Glm_5_1Params, "__pydantic_config__", _PARAMS_CONFIG)
|
|
96
96
|
|
|
97
|
+
Glm_5_2Params = TypedDict(
|
|
98
|
+
"Glm_5_2Params",
|
|
99
|
+
{
|
|
100
|
+
"max_completion_tokens": Annotated[int, Field(ge=1)],
|
|
101
|
+
"temperature": Annotated[float, Field(ge=0, le=1.9)],
|
|
102
|
+
"top_p": Annotated[float, Field(ge=0, le=1)],
|
|
103
|
+
"extra_body.top_k": Annotated[int, Field(ge=0)],
|
|
104
|
+
"extra_body.enable_thinking": bool,
|
|
105
|
+
"extra_body.thinking_budget": Annotated[int, Field(ge=1)],
|
|
106
|
+
},
|
|
107
|
+
total=False,
|
|
108
|
+
)
|
|
109
|
+
setattr(Glm_5_2Params, "__pydantic_config__", _PARAMS_CONFIG)
|
|
110
|
+
|
|
97
111
|
Kimi_K2_7_CodeParams = TypedDict(
|
|
98
112
|
"Kimi_K2_7_CodeParams",
|
|
99
113
|
{
|
|
@@ -523,6 +537,7 @@ __all__ = [
|
|
|
523
537
|
"Deepseek_V4_ProParams",
|
|
524
538
|
"Deepseek_V4_Pro_0813Params",
|
|
525
539
|
"Glm_5_1Params",
|
|
540
|
+
"Glm_5_2Params",
|
|
526
541
|
"Kimi_K2_7_CodeParams",
|
|
527
542
|
"Qwen_FlashParams",
|
|
528
543
|
"Qwen_MaxParams",
|
|
@@ -14,8 +14,12 @@ Kimi_K2_5Params = TypedDict(
|
|
|
14
14
|
"Kimi_K2_5Params",
|
|
15
15
|
{
|
|
16
16
|
"max_completion_tokens": Annotated[int, Field(ge=1)],
|
|
17
|
+
"stop": str,
|
|
17
18
|
"thinking.type": Literal["enabled", "disabled"],
|
|
18
19
|
"response_format.type": Literal["text", "json_object"],
|
|
20
|
+
"tool_choice": Literal["auto", "none"],
|
|
21
|
+
"logprobs": bool,
|
|
22
|
+
"top_logprobs": Annotated[int, Field(ge=0, le=20)],
|
|
19
23
|
},
|
|
20
24
|
total=False,
|
|
21
25
|
)
|
|
@@ -25,8 +29,13 @@ Kimi_K2_6Params = TypedDict(
|
|
|
25
29
|
"Kimi_K2_6Params",
|
|
26
30
|
{
|
|
27
31
|
"max_completion_tokens": Annotated[int, Field(ge=1)],
|
|
32
|
+
"stop": str,
|
|
28
33
|
"thinking.type": Literal["enabled", "disabled"],
|
|
34
|
+
"thinking.keep": Literal["all", None],
|
|
29
35
|
"response_format.type": Literal["text", "json_object"],
|
|
36
|
+
"tool_choice": Literal["auto", "none"],
|
|
37
|
+
"logprobs": bool,
|
|
38
|
+
"top_logprobs": Annotated[int, Field(ge=0, le=20)],
|
|
30
39
|
},
|
|
31
40
|
total=False,
|
|
32
41
|
)
|
|
@@ -37,6 +46,7 @@ Kimi_K2_6_SubscriptionParams = TypedDict(
|
|
|
37
46
|
{
|
|
38
47
|
"max_completion_tokens": Annotated[int, Field(ge=1)],
|
|
39
48
|
"thinking.type": Literal["enabled", "disabled"],
|
|
49
|
+
"thinking.keep": Literal["all", None],
|
|
40
50
|
"response_format.type": Literal["text", "json_object"],
|
|
41
51
|
},
|
|
42
52
|
total=False,
|
|
@@ -47,7 +57,12 @@ Kimi_K2_7_CodeParams = TypedDict(
|
|
|
47
57
|
"Kimi_K2_7_CodeParams",
|
|
48
58
|
{
|
|
49
59
|
"max_completion_tokens": Annotated[int, Field(ge=1)],
|
|
60
|
+
"stop": str,
|
|
61
|
+
"thinking.type": Literal["enabled"],
|
|
50
62
|
"response_format.type": Literal["text", "json_object", "json_schema"],
|
|
63
|
+
"tool_choice": Literal["auto", "none"],
|
|
64
|
+
"logprobs": bool,
|
|
65
|
+
"top_logprobs": Annotated[int, Field(ge=0, le=20)],
|
|
51
66
|
},
|
|
52
67
|
total=False,
|
|
53
68
|
)
|
|
@@ -57,7 +72,12 @@ Kimi_K2_7_Code_HighspeedParams = TypedDict(
|
|
|
57
72
|
"Kimi_K2_7_Code_HighspeedParams",
|
|
58
73
|
{
|
|
59
74
|
"max_completion_tokens": Annotated[int, Field(ge=1)],
|
|
75
|
+
"stop": str,
|
|
76
|
+
"thinking.type": Literal["enabled"],
|
|
60
77
|
"response_format.type": Literal["text", "json_object", "json_schema"],
|
|
78
|
+
"tool_choice": Literal["auto", "none"],
|
|
79
|
+
"logprobs": bool,
|
|
80
|
+
"top_logprobs": Annotated[int, Field(ge=0, le=20)],
|
|
61
81
|
},
|
|
62
82
|
total=False,
|
|
63
83
|
)
|
|
@@ -67,6 +87,7 @@ Kimi_K2_7_Code_Highspeed_SubscriptionParams = TypedDict(
|
|
|
67
87
|
"Kimi_K2_7_Code_Highspeed_SubscriptionParams",
|
|
68
88
|
{
|
|
69
89
|
"max_completion_tokens": Annotated[int, Field(ge=1)],
|
|
90
|
+
"thinking.type": Literal["enabled"],
|
|
70
91
|
"response_format.type": Literal["text", "json_object"],
|
|
71
92
|
},
|
|
72
93
|
total=False,
|
|
@@ -77,6 +98,7 @@ Kimi_K2_7_Code_SubscriptionParams = TypedDict(
|
|
|
77
98
|
"Kimi_K2_7_Code_SubscriptionParams",
|
|
78
99
|
{
|
|
79
100
|
"max_completion_tokens": Annotated[int, Field(ge=1)],
|
|
101
|
+
"thinking.type": Literal["enabled"],
|
|
80
102
|
"response_format.type": Literal["text", "json_object"],
|
|
81
103
|
},
|
|
82
104
|
total=False,
|
|
@@ -87,8 +109,11 @@ Kimi_K3Params = TypedDict(
|
|
|
87
109
|
"Kimi_K3Params",
|
|
88
110
|
{
|
|
89
111
|
"max_completion_tokens": Annotated[int, Field(ge=1)],
|
|
112
|
+
"stop": str,
|
|
90
113
|
"thinking.type": Literal["enabled", "disabled"],
|
|
114
|
+
"reasoning_effort": Literal["low", "high", "max"],
|
|
91
115
|
"response_format.type": Literal["text", "json_object", "json_schema"],
|
|
116
|
+
"tool_choice": Literal["auto", "none", "required"],
|
|
92
117
|
},
|
|
93
118
|
total=False,
|
|
94
119
|
)
|
|
@@ -98,12 +123,16 @@ Moonshot_V1_128kParams = TypedDict(
|
|
|
98
123
|
"Moonshot_V1_128kParams",
|
|
99
124
|
{
|
|
100
125
|
"max_completion_tokens": Annotated[int, Field(ge=1)],
|
|
126
|
+
"stop": str,
|
|
101
127
|
"temperature": Annotated[float, Field(ge=0, le=1)],
|
|
102
128
|
"top_p": Annotated[float, Field(ge=0, le=1)],
|
|
103
129
|
"n": Annotated[int, Field(ge=1, le=5)],
|
|
104
130
|
"presence_penalty": Annotated[float, Field(ge=-2, le=2)],
|
|
105
131
|
"frequency_penalty": Annotated[float, Field(ge=-2, le=2)],
|
|
106
132
|
"response_format.type": Literal["text", "json_object"],
|
|
133
|
+
"tool_choice": Literal["auto", "none", "required"],
|
|
134
|
+
"logprobs": bool,
|
|
135
|
+
"top_logprobs": Annotated[int, Field(ge=0, le=20)],
|
|
107
136
|
},
|
|
108
137
|
total=False,
|
|
109
138
|
)
|
|
@@ -113,12 +142,16 @@ Moonshot_V1_32kParams = TypedDict(
|
|
|
113
142
|
"Moonshot_V1_32kParams",
|
|
114
143
|
{
|
|
115
144
|
"max_completion_tokens": Annotated[int, Field(ge=1)],
|
|
145
|
+
"stop": str,
|
|
116
146
|
"temperature": Annotated[float, Field(ge=0, le=1)],
|
|
117
147
|
"top_p": Annotated[float, Field(ge=0, le=1)],
|
|
118
148
|
"n": Annotated[int, Field(ge=1, le=5)],
|
|
119
149
|
"presence_penalty": Annotated[float, Field(ge=-2, le=2)],
|
|
120
150
|
"frequency_penalty": Annotated[float, Field(ge=-2, le=2)],
|
|
121
151
|
"response_format.type": Literal["text", "json_object"],
|
|
152
|
+
"tool_choice": Literal["auto", "none", "required"],
|
|
153
|
+
"logprobs": bool,
|
|
154
|
+
"top_logprobs": Annotated[int, Field(ge=0, le=20)],
|
|
122
155
|
},
|
|
123
156
|
total=False,
|
|
124
157
|
)
|
|
@@ -128,12 +161,16 @@ Moonshot_V1_8kParams = TypedDict(
|
|
|
128
161
|
"Moonshot_V1_8kParams",
|
|
129
162
|
{
|
|
130
163
|
"max_completion_tokens": Annotated[int, Field(ge=1)],
|
|
164
|
+
"stop": str,
|
|
131
165
|
"temperature": Annotated[float, Field(ge=0, le=1)],
|
|
132
166
|
"top_p": Annotated[float, Field(ge=0, le=1)],
|
|
133
167
|
"n": Annotated[int, Field(ge=1, le=5)],
|
|
134
168
|
"presence_penalty": Annotated[float, Field(ge=-2, le=2)],
|
|
135
169
|
"frequency_penalty": Annotated[float, Field(ge=-2, le=2)],
|
|
136
170
|
"response_format.type": Literal["text", "json_object"],
|
|
171
|
+
"tool_choice": Literal["auto", "none", "required"],
|
|
172
|
+
"logprobs": bool,
|
|
173
|
+
"top_logprobs": Annotated[int, Field(ge=0, le=20)],
|
|
137
174
|
},
|
|
138
175
|
total=False,
|
|
139
176
|
)
|
|
File without changes
|
|
File without changes
|
|
File without changes
|
|
File without changes
|
|
File without changes
|
|
File without changes
|
|
File without changes
|
|
File without changes
|
|
File without changes
|
|
File without changes
|
|
File without changes
|
|
File without changes
|
|
File without changes
|
|
File without changes
|
|
File without changes
|
|
File without changes
|
|
File without changes
|
|
File without changes
|
|
File without changes
|
|
File without changes
|
|
File without changes
|
|
File without changes
|
|
File without changes
|
|
File without changes
|
|
File without changes
|
|
File without changes
|