-
`RuntimeError: CUDA error: CUBLAS_STATUS_INVALID_VALUE when calling `cublasSgemmStridedBatched( handle, opa, opb, m, n, k, &alpha, a, lda, stridea, b, ldb, strideb, &beta, c, ldc, stridec, num_batches…
-
Hello, thank you for your contribution to **`twin-offload`**. When I tried to run `ds_pretrain_gpt_2.7B.sh` at Megatron-Deepspeed with the latest parameter `"offload_optimizer":"ratio"`, I tried to se…
-
### Your current environment
```text
The output of `python collect_env.py`
```
Collecting environment information...
PyTorch version: 2.2.1+cu121
Is debug build: False
CUDA used to build PyTorc…
-
### Describe the bug
Seems like Tensor.tolist() is missing.
### Is there an existing issue for this?
- [X] I have searched the existing issues
### Reproduction
I tried to train a LoRA on Ouroboro…
Xabab updated
9 months ago
-
**Describe the bug**
cache[f"blocks.{x}.hook_resid_pre"] doesn't match hidden states (or only up to a set decimal place).
Hidden states is from transformer's model(tokens, output_hidden_states=Tru…
-
Hello
I tried converting GPT Neo-X from HF (https://huggingface.co/docs/transformers/model_doc/gpt_neox) to cTranslate2. I note this model was just recently supported.
I encountered an out of m…
-
### System Info
```Shell
- `Accelerate` version: 0.21.0.dev0
- Platform: Linux-5.15.109+-x86_64-with-glibc2.35
- Python version: 3.10.10
- Numpy version: 1.23.5
- PyTorch version (GPU?): 2.0.0 (T…
-
**Describe the bug**
**fused_kernels gives an error during installation**
**Console output:**
running install
running bdist_egg
running egg_info
writing fused_kernels.egg-info/PKG-INFO
writing …
-
### Feature request
Current [text-generation](https://github.com/huggingface/optimum-habana/tree/main/examples/text-generation) only support bloom & bloomz with deepspeed, but not support other gener…
-
Currently, n_ctx is locked to 2048, but with people starting to experiment with ALiBi models (BluemoonRP, MTP whenever that gets sorted out properly) and RedPajamas talking about hyena and StableLM ai…