Commit graph

29 commits

Author SHA1 Message Date
Yagil Burowski
c038696aa8
Add LM Studio support to the Assistant (#23097)
#### Release Notes:

- Added support for [LM Studio](https://lmstudio.ai/) to the Assistant.

#### Quick demo:


https://github.com/user-attachments/assets/af58fc13-1abc-4898-9747-3511016da86a

#### Future enhancements:
- wire up tool calling (new in [LM Studio
0.3.6](https://lmstudio.ai/blog/lmstudio-v0.3.6))

---------

Co-authored-by: Marshall Bowers <elliott.codes@gmail.com>
2025-01-14 20:41:58 +00:00
Max Brunsfeld
e309fbda2a
Add a slash command for automatically retrieving relevant context (#17972)
* [x] put this slash command behind a feature flag until we release
embedding access to the general population
* [x] choose a name for this slash command and name the rust module to
match

Release Notes:

- N/A

---------

Co-authored-by: Jason <jason@zed.dev>
Co-authored-by: Richard <richard@zed.dev>
Co-authored-by: Jason Mancuso <7891333+jvmncs@users.noreply.github.com>
Co-authored-by: Richard Feldman <oss@rtfeldman.com>
2024-09-20 18:09:18 -04:00
Richard Feldman
91ffa02e2c
/auto (#16696)
Add `/auto` behind a feature flag that's disabled for now, even for
staff.

We've decided on a different design for context inference, but there are
parts of /auto that will be useful for that, so we want them in the code
base even if they're unused for now.

Release Notes:

- N/A

---------

Co-authored-by: Antonio Scandurra <me@as-cii.com>
Co-authored-by: Marshall Bowers <elliott.codes@gmail.com>
2024-09-13 13:17:49 -04:00
Kyle Kelley
49371b44cb
Semantic Index (#10329)
This introduces semantic indexing in Zed based on chunking text from
files in the developer's workspace and creating vector embeddings using
an embedding model. As part of this, we've created an embeddings
provider trait that allows us to work with OpenAI, a local Ollama model,
or a Zed hosted embedding.

The semantic index is built by breaking down text for known
(programming) languages into manageable chunks that are smaller than the
max token size. Each chunk is then fed to a language model to create a
high dimensional vector which is then normalized to a unit vector to
allow fast comparison with other vectors with a simple dot product.
Alongside the vector, we store the path of the file and the range within
the document where the vector was sourced from.

Zed will soon grok contextual similarity across different text snippets,
allowing for natural language search beyond keyword matching. This is
being put together both for human-based search as well as providing
results to Large Language Models to allow them to refine how they help
developers.

Remaining todo:

* [x] Change `provider` to `model` within the zed hosted embeddings
database (as its currently a combo of the provider and the model in one
name)


Release Notes:

- N/A

---------

Co-authored-by: Nathan Sobo <nathan@zed.dev>
Co-authored-by: Antonio Scandurra <me@as-cii.com>
Co-authored-by: Conrad Irwin <conrad@zed.dev>
Co-authored-by: Marshall Bowers <elliott.codes@gmail.com>
Co-authored-by: Antonio <antonio@zed.dev>
2024-04-12 11:40:59 -06:00
KCaverly
68c37ca2a4 move embedding provider to ai crate 2023-09-22 09:33:59 -04:00
KCaverly
7b63369df2 move api authentication to embedding provider 2023-09-21 14:00:00 -04:00
Antonio Scandurra
ae85a520f2 Refactor semantic searching of modified buffers 2023-09-15 12:12:20 +02:00
KCaverly
e678c7d9ee swap
SystemTime for Instant throughout rate_limit_expiry tracking
2023-09-11 10:26:14 -04:00
KCaverly
37915ec4f2 updated notify to accomodate for updated countdown 2023-09-08 16:53:16 -04:00
KCaverly
bf43f93197 updated semantic_index reset status to leverage target reset system time as opposed to duration 2023-09-08 15:04:50 -04:00
KCaverly
a5ee8fc805 initial outline for rate limiting status updates 2023-09-08 12:35:15 -04:00
KCaverly
265d02a583 update request timeout for open ai embeddings 2023-09-06 15:09:46 -04:00
KCaverly
17237f748c update token_count for OpenAIEmbeddings to accomodate for truncation 2023-09-06 15:09:15 -04:00
Antonio Scandurra
3001a46f69 Reify Embedding/Sha1 structs that can be (de)serialized from SQL
Co-Authored-By: Kyle Caverly <kyle@zed.dev>
2023-08-31 17:55:43 +02:00
KCaverly
7d4d6c871b fix bug for truncation ensuring no valid inputs are sent to openai 2023-08-30 17:42:16 -04:00
KCaverly
76ce52df4e move queuing to embedding_queue functionality and update embedding provider to include trait items for max tokens per batch"
Co-authored-by: Max <max@zed.dev>
2023-08-30 16:01:28 -04:00
KCaverly
9781047156 move truncation to parsing step leveraging the EmbeddingProvider trait 2023-08-30 12:13:26 -04:00
KCaverly
76caea80f7 add should_truncate to embedding providers 2023-08-30 11:58:45 -04:00
KCaverly
e377ada1a9 added token count to documents during parsing 2023-08-30 11:05:46 -04:00
KCaverly
4f8b95cf0d add proper handling for open ai rate limit delays 2023-08-29 15:44:51 -04:00
KCaverly
3f9f742530 update rate limiting embeddings strategy to delay less 2023-08-24 11:45:52 +02:00
KCaverly
ced2b2aec3 reworked ProjectState to include additional context 2023-08-22 11:58:48 +02:00
KCaverly
3d89cd10a4 added sha1 encoding for each document 2023-08-21 16:35:57 +02:00
KCaverly
0e071919a0 parellelize embedding api calls 2023-07-18 16:09:44 -04:00
Max Brunsfeld
8d0614ce74 Populate project search results multi-buffer from semantic search
Co-authored-by: Kyle <kyle@zed.dev>
2023-07-18 11:44:58 -07:00
KCaverly
ed1b1a5ccd update logging for open ai embedding and remove redundant truncation 2023-07-18 11:00:21 -04:00
KCaverly
b9fdfd60f0 catch up with main 2023-07-18 10:26:28 -04:00
Max Brunsfeld
afc4c10ec1 Start work on exposing semantic search via project search view
Co-authored-by: Kyle <kyle@zed.dev>
2023-07-17 18:10:51 -07:00
KCaverly
8b42f5b1b3 rename vector_store crate to semantic_index 2023-07-17 17:06:10 -04:00
Renamed from crates/vector_store/src/embedding.rs (Browse further)