Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jamba - Skip 4d custom attention mask test #30826

Merged
merged 2 commits into from
May 15, 2024
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
1 change: 0 additions & 1 deletion src/transformers/models/jamba/modeling_jamba.py
Original file line number Diff line number Diff line change
Expand Up @@ -1261,7 +1261,6 @@ class JambaPreTrainedModel(PreTrainedModel):
_skip_keys_device_placement = "past_key_values"
_supports_flash_attn_2 = True
_supports_sdpa = True
_supports_cache_class = True

def _init_weights(self, module):
std = self.config.initializer_range
Expand Down
4 changes: 4 additions & 0 deletions tests/models/jamba/test_modeling_jamba.py
Original file line number Diff line number Diff line change
Expand Up @@ -502,6 +502,10 @@ def _prepare_model_kwargs(input_ids, attention_mask, signature):
# They should result in very similar logits
self.assertTrue(torch.allclose(next_logits_wo_padding, next_logits_with_padding, atol=3e-3))

@unittest.skip("Jamba has its own special cache type") # FIXME: @gante
def test_assisted_decoding_matches_greedy_search_0_random(self):
pass

@require_flash_attn
@require_torch_gpu
@require_bitsandbytes
Expand Down
Loading