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

[TT-Train] Fix text generation #17195

Merged
merged 1 commit into from
Jan 27, 2025
Merged

[TT-Train] Fix text generation #17195

merged 1 commit into from
Jan 27, 2025

Conversation

dmakoviichuk-tt
Copy link
Contributor

Problem description

Sampling from the padded logits may return a wrong tokens.

What's changed

Made sure you can never generate token > tokenizer vocab size.

Checklist

  • Post commit CI passes
  • Blackhole Post commit (if applicable)
  • Model regression CI testing passes (if applicable)
  • Device performance regression CI testing passes (if applicable)
  • (For models and ops writers) Full new models tests passes
  • New/Existing tests provide coverage for changes

@dmakoviichuk-tt dmakoviichuk-tt merged commit 28554c9 into main Jan 27, 2025
11 checks passed
@dmakoviichuk-tt dmakoviichuk-tt deleted the DM/fixed_generator branch January 27, 2025 22:59
williamlyTT pushed a commit that referenced this pull request Jan 30, 2025
### Problem description
Sampling from the padded logits may return a wrong tokens.

### What's changed
Made sure you can never generate token > tokenizer vocab size.

### Checklist
- [x] Post commit CI passes
- [x] Blackhole Post commit (if applicable)
- [x] Model regression CI testing passes (if applicable)
- [x] Device performance regression CI testing passes (if applicable)
- [x] **(For models and ops writers)** Full [new
models](https://github.com/tenstorrent/tt-metal/actions/workflows/full-new-models-suite.yaml)
tests passes
- [x] New/Existing tests provide coverage for changes
yieldthought pushed a commit that referenced this pull request Jan 31, 2025
### Problem description
Sampling from the padded logits may return a wrong tokens.

### What's changed
Made sure you can never generate token > tokenizer vocab size.

### Checklist
- [x] Post commit CI passes
- [x] Blackhole Post commit (if applicable)
- [x] Model regression CI testing passes (if applicable)
- [x] Device performance regression CI testing passes (if applicable)
- [x] **(For models and ops writers)** Full [new
models](https://github.com/tenstorrent/tt-metal/actions/workflows/full-new-models-suite.yaml)
tests passes
- [x] New/Existing tests provide coverage for changes
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants