Skip to content

Fix README: correct token IDs for LLaDA tokenizer #2

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

Open
usovmv wants to merge 1 commit into
base: main
Choose a base branch
from
Open

Fix README: correct token IDs for LLaDA tokenizer #2

Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
11 changes: 7 additions & 4 deletions README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -108,11 +108,14 @@ cmake --build build -j$(nproc)
```bash
# Pre-tokenize your prompt (diffuse-cpp does not include a tokenizer)
# Use transformers or tiktoken to get token IDs as a comma-separated list
# CRITICAL: Use the correct tokenizer for your model.
# LLaDA → GSAI-ML/LLaDA-8B-Instruct
# Dream → Dream-org/Dream-v0-Instruct-7B

# Generate text (example with token IDs)
# Generate text (example for LLaDA with "What is the capital of France?")
./build/diffuse-cli \
-m model.gguf \
--tokens "128000,3923,374,279,6864,315,9822,30" \
--tokens "2372,341,268,7706,300,11406,30" \
-n 256 \
-s 16 \
-t 12 \
Expand All @@ -123,7 +126,7 @@ cmake --build build -j$(nproc)

```python
from transformers import AutoTokenizer
tokenizer = AutoTokenizer.from_pretrained("LLaDA-8B-Instruct")
tokenizer = AutoTokenizer.from_pretrained("GSAI-ML/LLaDA-8B-Instruct", trust_remote_code=True)
tokens = tokenizer.encode("What is the capital of France?")
print(",".join(map(str, tokens)))
```
Expand Down Expand Up @@ -272,7 +275,7 @@ params.remasking = diffuse_remasking::ENTROPY_EXIT;
params.entropy_threshold = 1.5f;

// Generate
std::vector<int32_t> prompt_tokens = {128000, 3923, 374};
std::vector<int32_t> prompt_tokens = {2372, 341, 268, 7706, 300, 11406, 30};
auto output = diffuse_generate(ctx, prompt_tokens, 256, params);

// Cleanup
Expand Down