Commit Graph

27 Commits

Author SHA1 Message Date
Max Brunsfeld
1b1070e0f7
Add tracing needed for LLM rate limit dashboards (#16388)
Release Notes:

- N/A

---------

Co-authored-by: Marshall <marshall@zed.dev>
2024-08-16 17:52:31 -04:00
Nathan Sobo
907d76208d
Allow display name of custom Anthropic models to be customized (#16376)
Also added some docs for our settings.

Release Notes:

- N/A
2024-08-16 14:02:37 -06:00
Roy Williams
b4f5f5024e
Support 8192 output tokens for Claude Sonnet 3.5 (#16358)
Release Notes:

- Added support for 8192 output tokens from Claude Sonnet 3.5
(https://x.com/alexalbert__/status/1812921642143900036)
2024-08-16 11:47:39 -04:00
Roy Williams
46fb917e02
Implement Anthropic prompt caching (#16274)
Release Notes:

- Adds support for Prompt Caching in Anthropic. For models that support
it this can dramatically lower cost while improving performance.
2024-08-15 22:21:06 -05:00
Max Brunsfeld
4c390b82fb
Make LanguageModel::use_any_tool return a stream of chunks (#16262)
This PR is a refactor to pave the way for allowing the user to view and
edit workflow step resolutions. I've made tool calls work more like
normal streaming completions for all providers. The `use_any_tool`
method returns a stream of strings (which contain chunks of JSON). I've
also done some minor cleanup of language model providers in general,
removing the duplication around handling streaming responses.

Release Notes:

- N/A
2024-08-14 18:02:46 -07:00
Marshall Bowers
ebdb755fef
Surface upstream rate limits from Anthropic (#16118)
This PR makes it so hitting upstream rate limits from Anthropic result
in an HTTP 429 response instead of an HTTP 500.

To do this we need to surface structured errors out of the `anthropic`
crate.

Release Notes:

- N/A
2024-08-12 11:59:24 -04:00
Max Brunsfeld
33e120d964
Capture telemetry data on per-user monthly LLM spending (#16050)
Release Notes:

- N/A

---------

Co-authored-by: Marshall <marshall@zed.dev>
2024-08-09 16:38:37 -07:00
Bennet Bo Fenner
514b79e461
collab: Always use newest anthropic model version (#15978)
When Anthropic releases a new version of their models, Zed AI users
should always get access to the new version even when using an old
version of zed.

Co-Authored-By: Thorsten <thorsten@zed.dev>

Release Notes:

- N/A

Co-authored-by: Thorsten <thorsten@zed.dev>
2024-08-08 15:24:08 +02:00
Marshall Bowers
cf5f4dddf5
Authorize access to language model providers based on country (#15859)
This PR updates the LLM service to authorize access to language model
providers based on the requester's country.

We detect the country using Cloudflare's
[`CF-IPCountry`](https://developers.cloudflare.com/fundamentals/reference/http-request-headers/#cf-ipcountry)
header.

The country code is then checked against the list of supported countries
for the given LLM provider. Countries that are not supported will
receive an `HTTP 451: Unavailable For Legal Reasons` response.

Release Notes:

- N/A
2024-08-06 11:49:04 -04:00
Kirill Bulatov
9384f665bb
Properly extract errors from the Anthropic API (#15534)
Before, we missed "successful" responses with the API errors, now they
are properly shown in the assistant panel.


![image](https://github.com/user-attachments/assets/0c0936af-86c2-4def-9a58-25d5e0912b97)


Release Notes:

- N/A
2024-07-31 16:31:11 +03:00
Antonio Scandurra
99bc90a372
Allow customization of the model used for tool calling (#15479)
We also eliminate the `completion` crate and moved its logic into
`LanguageModelRegistry`.

Release Notes:

- N/A

---------

Co-authored-by: Nathan <nathan@zed.dev>
2024-07-30 16:18:53 +02:00
Antonio Scandurra
6e1f7c6e1d
Use tool calling instead of XML parsing to generate edit operations (#15385)
Release Notes:

- N/A

---------

Co-authored-by: Nathan <nathan@zed.dev>
2024-07-29 16:42:08 +02:00
Antonio Scandurra
d6bdaa8a91
Simplify LLM protocol (#15366)
In this pull request, we change the zed.dev protocol so that we pass the
raw JSON for the specified provider directly to our server. This avoids
the need to define a protobuf message that's a superset of all these
formats.

@bennetbo: We also changed the settings for available_models under
zed.dev to be a flat format, because the nesting seemed too confusing.
Can you help us upgrade the local provider configuration to be
consistent with this? We do whatever we need to do when parsing the
settings to make this simple for users, even if it's a bit more complex
on our end. We want to use versioning to avoid breaking existing users,
but need to keep making progress.

```json
"zed.dev": {
  "available_models": [
    {
      "provider": "anthropic",
        "name": "some-newly-released-model-we-havent-added",
        "max_tokens": 200000
      }
  ]
}
```

Release Notes:

- N/A

---------

Co-authored-by: Nathan <nathan@zed.dev>
2024-07-28 11:07:10 +02:00
Antonio Scandurra
70e895a8c7
Fix regression that caused Anthropic custom models to error (#15329)
/cc: @bennetbo 

Release Notes:

- N/A

Co-authored-by: Nathan <nathan@zed.dev>
2024-07-27 14:45:18 +02:00
Mikayla Maki
855048041d
Update http crate name (#15041)
Release Notes:

- N/A
2024-07-23 15:01:05 -07:00
Bennet Bo Fenner
d0f52e90e6
assistant: Overhaul provider infrastructure (#14929)
<img width="624" alt="image"
src="https://github.com/user-attachments/assets/f492b0bd-14c3-49e2-b2ff-dc78e52b0815">

- [x] Correctly set custom model token count
- [x] How to count tokens for Gemini models?
- [x] Feature flag zed.dev provider
- [x] Figure out how to configure custom models
- [ ] Update docs

Release Notes:

- Added support for quickly switching between multiple language model
providers in the assistant panel

---------

Co-authored-by: Antonio <antonio@zed.dev>
2024-07-23 19:48:41 +02:00
Antonio Scandurra
0155435142
Allow using a custom model when using zed.dev (#14933)
Release Notes:

- N/A
2024-07-22 12:25:53 +02:00
Sean Billig
d3b3e072a7
Make Claude 3.5 the default Anthropic model (#13324)
Release Notes:

- N/A

Co-authored-by: Antonio Scandurra <me@as-cii.com>
2024-06-21 18:47:38 +02:00
Antonio Scandurra
4072ad2858
Add support for Claude 3.5 Sonnet (#13371)
Release Notes:

- Added support for Claude 3.5 Sonnet.
2024-06-21 18:32:26 +02:00
Antonio Scandurra
6ff01b17ca
Improve model selection in the assistant (#12472)
https://github.com/zed-industries/zed/assets/482957/3b017850-b7b6-457a-9b2f-324d5533442e


Release Notes:

- Improved the UX for selecting a model in the assistant panel. You can
now switch model using just the keyboard by pressing `alt-m`. Also, when
switching models via the UI, settings will now be updated automatically.
2024-05-30 12:36:07 +02:00
Antonio Scandurra
de09409f01
Sanitize messages before sending them to Anthropic (#11810)
Release Notes:

- N/A

Co-authored-by: Nathan <nathan@zed.dev>
Co-authored-by: David <davidsp@anthropic.com>
2024-05-14 17:47:33 +02:00
Antonio Scandurra
5944caaa90
Add support for interacting with Claude in the assistant panel (#11798)
Release Notes:

- Added support for interacting with Claude in the assistant panel. You
can enable it by adding the following to your `settings.json`:

    ```json
    "assistant": {
        "version": "1",
        "provider": {
            "name": "anthropic"
        }
    }
    ```
2024-05-14 15:57:52 +02:00
Conrad Irwin
5515ba6043
Extract http from util (#11680)
This avoids the CLI linking libssl etc...

Release Notes:

- N/A
2024-05-10 15:50:20 -06:00
Kyle Kelley
6563330239
Supermaven (#10788)
Adds a supermaven provider for completions. There are various other
refactors amidst this branch, primarily to make copilot no longer a
dependency of project as well as show LSP Logs for global LSPs like
copilot properly.

This feature is not enabled by default. We're going to seek to refine it
in the coming weeks.

Release Notes:

- N/A

---------

Co-authored-by: Antonio Scandurra <me@as-cii.com>
Co-authored-by: Nathan Sobo <nathan@zed.dev>
Co-authored-by: Max <max@zed.dev>
Co-authored-by: Max Brunsfeld <maxbrunsfeld@gmail.com>
2024-05-03 12:50:42 -07:00
Marshall Bowers
1fbc04104c Move lints section to the top of Cargo.toml, to match the others 2024-04-18 15:53:48 -04:00
Kirill Bulatov
5602593089
Check license generation for every PR to avoid license-less crate additions (#10033)
Also fix `anthropic` crate and make it AGPL-licensed, as it's used in
the AGPL-licensed collab part only.

Release Notes:

- N/A
2024-04-01 12:16:16 +03:00
Nathan Sobo
9b673089db
Enable Claude 3 models to be used via the Zed server if "language-models" feature flag is enabled for user (#10015)
Release Notes:

- N/A
2024-03-31 15:57:57 -06:00