This PR updates multiple default arguments in clients and servers:
1. **The client defaults to `torch_dtype=torch.float32` instead of `torch_dtype="auto"`.**
The old default was to load weights in the dtype they are saved in (usually bfloat16/float16), which caused issues when the client was run on CPU (the default unless you call `.cuda()`). Specifically, bfloat16 is slow on most CPUs (unless a CPU supports AVX512) and float16 can't be run natively and leads to an exception. This default was a legacy of the earliest Petals versions designed to run BLOOM - its embeddings were so big that they didn't fit into RAM in float32 (e.g., in Colab). The newer models don't have this issue.
In contrast, the new default leads to good speed on all CPUs and is consistent with PyTorch and HF Transformers. Also, the client now shows "bfloat16 on non-AVX512 CPU" in all cases (previously this warning was shown only if the machine has enough RAM to fit float32 weights, which could hide the crucial reason of inference being slow).
**Note:** This change is backward-incompatible, so we have to increase at least the minor package version (2.2.0 -> 2.3.0.dev0).
2. **The server uses 2x smaller `--attn_cache_tokens`.**
The old default led to loading 39 (out of 80) or 78 (out of 80) blocks for popular models on some GPU types, which visibly slowed down inference due to an excess network hop. It was also leaving too much cache, so that inference slowed down much before the cache is used.
The new default leads to more efficient block layouts and makes the inference routing algorithm choose alternative paths through other servers when a particular server already has enough active inference sessions (= its cache is full).
3. **The client's max number of retries can be limited by the `PETALS_MAX_RETRIES` env var.**
This is to limit `ClientConfig.max_retries` in tests, so we see tracebacks instead of retrying indefinitely in case of errors.
This PR fixes gaps in the DHT server info caused by unavailable DHT keys. Now, one DHT key is enough to get info about all blocks hosted by a server - so we'll see info until all keys are unavailable.
Also, this PR refactors `petals.client.routing` and `petals.server.block_selection` modules to use the common `compute_spans()` function (defined in `petals.utils.dht`) and `RemoteSpanInfo` class (defined in `petals.data_structures`).
This PR attempts to optimize the inference of Falcon models in the single-token setup by reducing the majority of Python overhead and making several assumptions about the setup. Specifically,
* Layer normalization, QKV projection (with splitting) and rotary embeddings are executed through CUDA graphs, which reduces most overhead related to small kernel launche
* If no sin/cos tensors are cached by the rotary embedding layer, we cache them for 8192 tokens (INFERENCE_MAX_LENGTH) during the first forward pass. In general, it should be beneficial to always run a max-length sequence before starting a block, but this is a question for another PR
The PR also adds a small test to ensure that the results (without quantization) of the block before and after quantization indeed match.
Lastly, the pull request makes the backward pass work (as discussed in https://github.com/bigscience-workshop/petals/pull/499) by making cached sin/cos for RotaryEmbedding into buffers and disabling the inference mode during their creation.
This PR adds:
- Support for models based on `transformers.FalconModel` (the in-library format for Falcon). Tested on Falcon-40B.
- CI tests for Falcon-RW-1B.
- `--throughput dry_run` option to evaluate throughput and exit right away (implemented by @mryab).
Limitations:
- Backward pass support is broken for now, will be fixed in #500.
Co-authored-by: Max Ryabinin <mryabinin0@gmail.com>
This PR creates an index of models hosted in the swarm - it is useful to know which custom models users run and display them at https://health.petals.dev as "not officially supported" models.
This PR makes both clients and servers work on macOS. Specifically, it:
- Follows https://github.com/learning-at-home/hivemind/pull/586 to run a macOS-compatible `p2pd` binary (both x86-64 and ARM64 are supported)
- Fixes forking issues and tests on macOS, Python 3.10+
- Introduces basic support for serving model blocks on Apple M1/M2 GPUs (torch.mps)
- Increases max number of open files by default (it's not enough on Linux and is really small on macOS)
- rpc_inference: server will now accept allocation timeout from user, defaults to no timeout
- bugfix: inference timeout is now measured from the moment the request is received
- previously, you would have to wait for your timeout plus the time it takes to sort through the queue (other users' timeout)
- now, you get AllocationFailed if you had to wait for over (timeout) seconds - regardless of other users
- a request for inference with no timeout will now fail instantly if there is not enough memory available
- dtype number of bytes is now correctly determined for int, bool & other types
---------
Co-authored-by: Your Name <you@example.com>
Co-authored-by: Alexander Borzunov <borzunov.alexander@gmail.com>
Co-authored-by: Aleksandr Borzunov <hxrussia@gmail.com>
Before:
```python
Aug 23 23:51:31.394 [INFO] Loaded Maykeye/TinyLLama-v0 block 0, _IncompatibleKeys(missing_keys=[], unexpected_keys=['self_attn.rotary_emb.inv_freq'])
```
After:
```python
Aug 23 23:51:31.394 [INFO] Loaded Maykeye/TinyLLama-v0 block 0
```
Hiding this since the excess keys in Llama-based models are okay since the latest transformers release.
Peft 0.5 recently released and broke some compatilibities. This PR temporarily requires petals to use the previous stable version of peft while we work on 0.5.0 support.
This PR drops custom generation codes and introduces compatibility with `transformers.GenerationMixin` instead. This includes support for more sampling options (`top_p`, `top_k`, `repetition_penalty` requested in #460) and beam search - all that is now identical to running model with transformers locally.
Most features (excluding beam search and other rarely used stuff) are also compatible with resuming existing sessions.
### Breaking changes
If `.generate()` or forward passes are being run inside an `.inference_session()` context, they now use the opened session by default. So, these snippets are now equivalent:
```python
# Using default session
with model.inference_session(max_length=100):
output_ids = model.generate(input_ids, max_new_tokens=3)
# Explicitly specifying a session
with model.inference_session(max_length=100) as sess:
output_ids = model.generate(input_ids, max_new_tokens=3, session=sess)
```
Earlier, the 1st snippet was creating a new session, which is not what most people expected (= such code was most likely to introduce a bug, which is now fixed).
Right now, long inference requests may occupy Runtime for a few seconds without giving it away to process short (most latency-sensitive requests). This PR fixes it by disallowing the merged pool for long requests and prioritizing the short ones.
This PR extends CI to:
1. Test Llama code using [TinyLlama-v0](https://huggingface.co/Maykeye/TinyLLama-v0).
2. Test rebalancing (sets up a situation where the 1st server needs to change its original position).
3. Check if benchmark scripts run (in case someone breaks its code). Note that the benchmark results are meaningless here (since they're measured on a tiny swarm of CPU servers, with low `--n_steps`).
4. Test `petals.cli.run_dht`.
5. Increase swap space and watch free RAM (a common issue is that actions are cancelled without explanation if there's not enough RAM - so it's a useful reminder + debug tool).
6. Fix flapping tests for bloom-560m by increasing tolerance.
Other minor changes: fix `--help` messages to show defaults, fix docs, tune rebalancing constants.
This prohibits passing `--initial_peers` without arguments, since it's likely to be a side-effect from `--initial_peers $INITIAL_PEERS` with the env var not set.
Users should use `--new_swarm` for that, as explained in the private swarm tutorial.
We choose longer servers to minimize the number of hops but leave some randomization to distribute the load. We also exclude servers known to be unreachable.
