Commit Graph

25 Commits

Author SHA1 Message Date
Nicolas Patry fb2f74e2b9
Refactor dead code - Removing all `flash_xxx.py` files. (#2166)
* Refactor dead code.

* First working step.

* Remove a lot of duplicated code.

* More dead code.

* More cleanup.

* Fix Santacoder test.

* Fixing the simple tests.

* Fixing sharding.

* Fixes for VLM.

* Fixing santacoder (num_kv_heads hardcoded).

* Removing more dead code.

* Fixing `config.n_head`.

* Stopping earlier because of `<end_of_utterance>` in idefics2.

* Addresses comments.

* Removing the dead code.

* Fuse back mistral into FlashCausalLM.

* Finish removal.

* Fixing docs + causal_lm `batch_class`.

* Fixing docs + causal.lm.

* Add default to Gemma Causality.

* Default value for gemma/gemma2.

* Wrong default.
2024-07-05 10:29:56 +02:00
Daniël de Kok bf3c813782 server: use chunked inputs
The router will now send the input as chunks besides as a single
string. This change modifies the server to process chunked input
rather than strings. This also allows us to remove the image
extraction code from the server.
2024-06-07 08:09:04 +02:00
OlivierDehaene 50b495f3d8
feat: add more latency metrics in forward (#1346) 2023-12-14 15:59:38 +01:00
OlivierDehaene 72ee382ded chore: formatting 2023-12-11 14:49:52 +01:00
Nicolas Patry 9ecfa16b12
Speculative (#1308) 2023-12-11 12:46:30 +01:00
Nicolas Patry abd58ff82c
feat(server): Rework model loading (#344)
# What does this PR do?

Reworked the loading logic. Idea is to use cleaner loading code:

- Remove need for `no_init_weights`
- Remove all weird `bnb_linear` and `load_weights` and
`post_load_weights`.

New code layout:

- New class `Weights` in charge of handling loading the weights from
multiple files into appropiate tensors (potentially sharded)
- TP layers now are "shells", they contain the code to know what kind of
sharding we need + eventual `all_reduce`. They do not inherit from
linear, but they contain some kind of Linear instead
- the contained linear can be either FastLinear, BnbLinear or GPTq
Linear next.
- All modeling code is explictly made for sharding, process group is
just no-ops for non sharded code (removes a lot of test cases)

![Screenshot from 2023-05-19
23-19-59](https://github.com/huggingface/text-generation-inference/assets/204321/9a802654-74a3-488c-87a8-073743a6143f)

---------

Co-authored-by: Ubuntu <ubuntu@ip-172-31-41-161.taildb5d.ts.net>
Co-authored-by: Ubuntu <ubuntu@ip-172-31-41-161.ec2.internal>
Co-authored-by: OlivierDehaene <olivier@huggingface.co>
Co-authored-by: OlivierDehaene <23298448+OlivierDehaene@users.noreply.github.com>
2023-06-08 14:51:52 +02:00
OlivierDehaene 895c5f1562
feat(server): only compute prefill logprobs when asked (#406)
Close #288
2023-06-02 17:12:30 +02:00
OlivierDehaene 62f91f78ac
feat(server): support vectorized warpers in flash causal lm (#317)
Co-authored-by: Joel Lamy-Poirier <joel.lamy-poirier@servicenow.com>
2023-05-26 12:30:27 +02:00
OlivierDehaene 218c9adaa5
feat: decrease IPC proto size (#367)
Closes #307 #308
2023-05-24 19:19:57 +02:00
OlivierDehaene ebc74d5666
feat(router): use number of tokens in batch as input for dynamic batching (#226)
Co-authored-by: Nick Hill <nickhill@us.ibm.com>
2023-04-24 17:59:00 +02:00
Nick Hill 4a7dd4085a
feat(server): reduce memory requirement (#214) 2023-04-24 14:15:42 +02:00
OlivierDehaene 709d8936f6
feat(router): drop requests when client closes the channel (#202) 2023-04-20 11:07:40 +02:00
OlivierDehaene 9987960062
feat(router): make router input validation optional (#164) 2023-04-09 20:22:27 +02:00
OlivierDehaene b49dbf2d88
fix(server): use server tokenizer as gt (#128) 2023-03-16 12:12:26 +01:00
OlivierDehaene 3fef90d50f
feat(clients): Python client (#103) 2023-03-07 18:52:22 +01:00
OlivierDehaene 44ce098c10
feat(server): pre-allocate max attention mask (#75) 2023-02-24 12:49:21 +01:00
OlivierDehaene 20c3c5940c
feat(router): refactor API and add openAPI schemas (#53) 2023-02-03 12:43:37 +01:00
OlivierDehaene b1482d9048
breaking(router): modify /generate API to only return generated text (#50)
@njhill, @yk FYI

generated_text was concatenated to the user prompt for legacy reason. We
want to remove this behaviour as we don't think it is useful and even
detrimonial to usability.

We also remove the unused Vec.
2023-02-02 15:02:04 +01:00
OlivierDehaene 017a2a8c2f
feat: Add token streaming using ServerSideEvents support (#41) 2023-01-31 17:04:00 +01:00
OlivierDehaene 4f9ac67cfa
Revert "feat: Add token streaming using ServerSideEvents support" (#40)
Reverts huggingface/text-generation-inference#36
2023-01-31 14:21:51 +01:00
OlivierDehaene 7fbfbb0dc5
feat: Add token streaming using ServerSideEvents support (#36)
Add token streaming using ServerSideEvents (SSE).

The signature of the SSE events is: 

```rust
struct Details {
    finish_reason: String,
    generated_tokens: u32,
    seed: Option<u64>,
}

struct StreamResponse {
    token: Token,
    generated_text: Option<String>,
    details: Option<Details>,
}

struct ErrorResponse {
    error: String,
}
```
2023-01-31 11:49:43 +01:00
OlivierDehaene 15511edc01
feat(server): Support SantaCoder (#26) 2023-01-20 12:24:39 +01:00
OlivierDehaene 32a253063d
feat: Return logprobs (#8) 2022-12-15 17:03:56 +01:00
OlivierDehaene 718096f695
feat: Support stop sequences (#7) 2022-12-12 18:25:22 +01:00
OlivierDehaene a2985036aa
feat(server): Add model tests (#6) 2022-12-08 18:49:33 +01:00