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.

Sign up for GitHub

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

Fixes MPS device errors from Tensor.type() when using generate_text_semantic and generate_coarse #27

Merged
merged 5 commits into from
May 14, 2023
Merged

Fixes MPS device errors from Tensor.type() when using generate_text_semantic and generate_coarse #27

Changes from all commits
Commits
Show all changes
5 commits
Select commit Hold shift + click to select a range
b21d8c6
Force type as pytorch Tensor.type() has missing support for MPS.
fiq May 3, 2023
89b362b
Fix to address pytorch Tensor.type() having missing support for MPS.
fiq May 3, 2023
c95350f
DRY: refactored mps supporting dtype.
fiq May 4, 2023
8660b7d
simplified fix for pytorch MPS .type() issue
fiq May 7, 2023
c84f71e
preserve device type in a temporary variable
fiq May 13, 2023
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
12 changes: 4 additions & 8 deletions bark/generation.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -339,7 +339,6 @@ def preload_models(
# Generation Functionality
####


def _tokenize(tokenizer, text):
return tokenizer.encode(text, add_special_tokens=False)

Expand All @@ -351,7 +350,6 @@ def _detokenize(tokenizer, enc_text):
def _normalize_whitespace(text):
return re.sub(r"\s+", " ", text).strip()


TEXT_ENCODING_OFFSET = 10_048
SEMANTIC_PAD_TOKEN = 10_000
TEXT_PAD_TOKEN = 129_595
Expand Down Expand Up @@ -453,8 +451,7 @@ def generate_text_semantic(
)
if top_p is not None:
# faster to convert to numpy
logits_device = relevant_logits.device
logits_dtype = relevant_logits.type()
original_device = relevant_logits.device
relevant_logits = relevant_logits.detach().cpu().type(torch.float32).numpy()
sorted_indices = np.argsort(relevant_logits)[::-1]
sorted_logits = relevant_logits[sorted_indices]
Expand All @@ -464,7 +461,7 @@ def generate_text_semantic(
sorted_indices_to_remove[0] = False
relevant_logits[sorted_indices[sorted_indices_to_remove]] = -np.inf
relevant_logits = torch.from_numpy(relevant_logits)
relevant_logits = relevant_logits.to(logits_device).type(logits_dtype)
relevant_logits = relevant_logits.to(original_device)
if top_k is not None:
v, _ = torch.topk(relevant_logits, min(top_k, relevant_logits.size(-1)))
relevant_logits[relevant_logits < v[-1]] = -float("Inf")
Expand Down Expand Up @@ -649,8 +646,7 @@ def generate_coarse(
relevant_logits = logits[0, 0, logit_start_idx:logit_end_idx]
if top_p is not None:
# faster to convert to numpy
logits_device = relevant_logits.device
logits_dtype = relevant_logits.type()
original_device = relevant_logits.device
relevant_logits = relevant_logits.detach().cpu().type(torch.float32).numpy()
sorted_indices = np.argsort(relevant_logits)[::-1]
sorted_logits = relevant_logits[sorted_indices]
Expand All @@ -660,7 +656,7 @@ def generate_coarse(
sorted_indices_to_remove[0] = False
relevant_logits[sorted_indices[sorted_indices_to_remove]] = -np.inf
relevant_logits = torch.from_numpy(relevant_logits)
relevant_logits = relevant_logits.to(logits_device).type(logits_dtype)
relevant_logits = relevant_logits.to(original_device)
if top_k is not None:
v, _ = torch.topk(relevant_logits, min(top_k, relevant_logits.size(-1)))
relevant_logits[relevant_logits < v[-1]] = -float("Inf")
Expand Down