de11fcddb1
Initial prototype of register_template_loaders plugin hook
...
Refs #809
2025-02-28 22:23:14 -08:00
b829cd92e0
Show async in list of features, closes #806
2025-02-28 06:44:09 -08:00
3a60290c82
llm logs --id-gt and --id-gte options, closes #801
2025-02-28 00:15:59 -08:00
48f67f4085
llm logs --data-ids flag, closes #800
2025-02-27 20:31:50 -08:00
1bebf8b34a
--schema t:template-name option, plus improved schema docs
...
Closes #799 , refs #788
2025-02-27 17:25:31 -08:00
b1fe2e9857
sort_keys=False in --save, closes #798
2025-02-27 16:51:43 -08:00
98cccd294a
llm models list --schemas option, closes #797
...
Also fixed bug where features showed even without --options, refs #796
2025-02-27 15:50:28 -08:00
4a7a1f19ed
Show features (including streaming) in llm models --options, closes #796
2025-02-27 15:44:38 -08:00
efe265137d
Only some OpenAI models support schemas, closes #794
2025-02-27 15:02:39 -08:00
6bec92fd78
Assign gpt-4.5 default alias, refs #795
2025-02-27 14:51:09 -08:00
801b08bf40
gpt-4.5-preview and gpt-4.5-preview-2025-02-27, refs #795
2025-02-27 12:25:04 -08:00
a1ea85ecbd
llm logs --schema-multi option
2025-02-27 11:20:23 -08:00
9a38021218
llm schemas dsl command, closes #793
...
Refs #790
2025-02-27 10:46:56 -08:00
eb2b243fdf
schema_dsl(..., multi=True) parameter, refs #790
2025-02-27 10:28:42 -08:00
8d32b71ef1
Renamed build_json_schema to schema_dsl
2025-02-27 10:22:29 -08:00
7e819c2ffa
Implemented --schema-multi, closes #791
2025-02-27 10:12:21 -08:00
e20bd45cb5
Fix for mypy and ruff, refs #790
2025-02-27 10:01:24 -08:00
321636e791
New schema DSL, closes #790
...
Plus made a start on schemas.md refs #788
2025-02-27 09:48:44 -08:00
c126b5d04c
llm schemas list --full option
...
Refs https://github.com/simonw/llm/issues/781#issuecomment-2688342157
2025-02-27 07:43:41 -08:00
4908cdfbd2
llm schemas show X command, refs #781
2025-02-27 07:39:36 -08:00
99a1adcece
Initial llm schemas list implementation, refs #781
2025-02-27 07:35:48 -08:00
a0845874ec
Schema template --save --schema support
...
* Don't hang on stdin if llm -t template-with-schema
* Docs on using schemas with templates
* Schema in template YAML file example
* Test for --save with --schema
Refs #778
2025-02-27 07:19:15 -08:00
f35ac31c21
llm logs --schema, --data, --data-array and --data-key options ( #785 )
...
* llm logs --schema option, refs #782
* --data and --data-array and --data-key options, refs #782
* Tests for llm logs --schema options, refs #785
* Also implemented --schema ID lookup, refs #780
* Using --data-key implies --data
* Docs for llm logs --schema and --data etc
2025-02-26 21:51:08 -08:00
6cb16a1d1a
Allow "reasoning" for extra-openai-models.yaml ( #766 )
...
* Allow "reasoning" for extra-openai-models.yaml
Currently you get an error when trying to use `-o reasoning_effort high` with a model that has been defined in `extra-openai-models.yaml`.
This allows a `reasoning` field.
* Mention reasoning: true in other OpenAI models docs
---------
Co-authored-by: Simon Willison <swillison@gmail.com >
2025-02-26 21:50:14 -08:00
9922d5bb6a
model.prompt(prompt= is now optional, closes #784
2025-02-26 19:31:38 -08:00
c58e3a2246
Use -- none -- for no prompt
...
Refs https://github.com/simonw/llm/issues/783#issuecomment-2686760812
2025-02-26 19:28:19 -08:00
3d4871f163
Improved log Markdown, closes #783
2025-02-26 19:25:25 -08:00
62c90dd472
llm prompt --schema X option and model.prompt(..., schema=) parameter ( #777 )
...
Refs #776
* Implemented new llm prompt --schema and model.prompt(schema=)
* Log schema to responses.schema_id and schemas table
* Include schema in llm logs Markdown output
* Test for schema=pydantic_model
* Initial --schema CLI documentation
* Python docs for schema=
* Advanced plugin docs on schemas
2025-02-26 16:58:28 -08:00
eda1f4f588
Add note about similarity function in "similar" command's doc ( #774 )
...
* note about similarity function in similar command doc
* Link to Wikipedia definition
---------
Co-authored-by: Simon Willison <swillison@gmail.com >
2025-02-26 10:07:10 -08:00
849c65fe9d
Upgrade to Pydantic v2 ( #775 )
...
* Upgrade to Pydantic v2
* Stop testing against Pydantic v1
Closes #520
2025-02-26 10:05:54 -08:00
7bf1ea665e
Made load_conversation() async_ aware, closes #742
2025-02-16 20:19:38 -08:00
24b250506b
Better __repr__ and __str__ for conversation and model
...
Inspired by work on #752
2025-02-16 15:45:08 -08:00
2066397aae
llm logs --prompts is now -s/--short - also supports --usage
...
Refs #736 , closes #756
2025-02-16 15:12:28 -08:00
6c6b100f3e
KeyModel and AsyncKeyModel classes for models that taken keys ( #753 )
...
* New KeyModel and AsyncKeyModel classes for models that taken keys - closes #744
* llm prompt --key now uses new mechanism, including for async
* use new key mechanism in llm chat command
* Python API tests for llm.KeyModel and llm.AsyncKeyModel
* Python API docs for for prompt(... key="")
* Mention await model.prompt() takes other parameters, reorg sections
* Better title for the model tutorial
* Docs on writing model plugins that take a key
2025-02-16 14:38:51 -08:00
b331b1e674
OpenAI model chatgpt-4o-latest, closes #752
2025-02-15 17:45:09 -08:00
31e900e9e1
llm aliases set -q option, refs #749
2025-02-13 15:49:47 -08:00
20c18a716d
-q multiple option for llm models and llm embed-models
...
Refs #748
2025-02-13 15:35:18 -08:00
9a1374b447
llm embed-multi --prepend option ( #746 )
...
* llm embed-multi --prepend option
Closes #745
2025-02-12 15:19:18 -08:00
41d64a8f12
llm logs --prompts option ( #737 )
...
Closes #736
2025-02-02 12:03:01 -08:00
deb8bc3b4f
Upgrade to black>=25.1.0
...
Refs https://github.com/simonw/llm/issues/728#issuecomment-2628348988
Refs https://github.com/psf/black/issues/4571#issuecomment-2628355450
2025-01-31 13:18:41 -08:00
965ad819f9
Fix for tests with pydantic<2, refs #728
2025-01-31 12:18:33 -08:00
eb0e1e761b
o3-mini and reasoning_effort option, refs #728
2025-01-31 12:14:02 -08:00
656d8fa3c4
--xl/--extract-last flag for prompt and log list commands ( #718 )
...
Closes #717
2025-01-24 10:52:46 -08:00
57d3baac42
Update embedding model names in docs, refs #654
...
Also ran Black.
2025-01-22 20:35:17 -08:00
6f7ea406bf
Register full embedding model names ( #654 )
...
Provide backward compatible aliases.
This makes available the same model names that ttok uses.
2025-01-22 20:14:03 -08:00
02e59a201e
Don't show default model for llm models -q, closes #710
2025-01-18 14:24:18 -08:00
f95dd55cda
Make it easier to debug CLI errors in pytest
...
Found this pattern while working on #709
2025-01-18 14:21:43 -08:00
e3c104b136
Show the default model when listing all available models. ( #688 )
2025-01-11 12:04:39 -08:00
73043ec406
Fixed mypy complaint
2025-01-10 16:05:29 -08:00
38a7366d8e
o1 cannot stream
...
https://github.com/simonw/llm/issues/676#issuecomment-2584932453
2025-01-10 16:03:09 -08:00
Simon Willison