https://github.com/mozzilemon/print_gguf
print_gguf.py is a simple utility to parse the header & tensor_infos of GGUF file.
https://github.com/mozzilemon/print_gguf
gguf llamacpp
Last synced: 8 months ago
JSON representation
print_gguf.py is a simple utility to parse the header & tensor_infos of GGUF file.
- Host: GitHub
- URL: https://github.com/mozzilemon/print_gguf
- Owner: mozzilemon
- License: mit
- Created: 2024-01-30T16:04:54.000Z (over 1 year ago)
- Default Branch: main
- Last Pushed: 2024-02-01T12:14:26.000Z (over 1 year ago)
- Last Synced: 2025-01-31T06:51:15.054Z (9 months ago)
- Topics: gguf, llamacpp
- Language: Python
- Homepage: https://github.com/mozzilemon
- Size: 7.81 KB
- Stars: 5
- Watchers: 1
- Forks: 1
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# print_gguf
print_gguf.py is a simple utility to parse the `header` & `tensor_infos` of GGUF file.
### How to use?
```python
python print_gguf.py {.gguf_file}
```
### Output Example)
```python
python print_gguf_info.py llama-v1-7b-q2k.gguf
magic = 0x46554747
version = 3
tensor_count = 291
metadata_kv_count = 16
general.architecture = llama
general.name = LLaMA
llama.context_length = 2048
llama.embedding_length = 4096
llama.block_count = 32
llama.feed_forward_length = 11008
llama.rope.dimension_count = 128
llama.attention.head_count = 32
llama.attention.head_count_kv = 32
llama.attention.layer_norm_rms_epsilon = 9.999999974752427e-07
general.file_type = 10
tokenizer.ggml.model = llama
tokenizer.ggml.tokens = ['', '', '', '<0x00>', '<0x01>', '<...
tokenizer.ggml.scores = [0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0...
tokenizer.ggml.token_type = [2, 3, 3, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6...
general.quantization_version = 2
name = token_embd.weight
n_dimensions = 2
shape = [4096, 32000]
ggml_type = GGML_TYPE_Q2_K
offset = 0
==================================================
name = output_norm.weight
n_dimensions = 1
shape = [4096]
ggml_type = GGML_TYPE_FP32
offset = 43008000
==================================================
name = output.weight
n_dimensions = 2
shape = [4096, 32000]
ggml_type = GGML_TYPE_Q6_K
offset = 43024384
==================================================
name = blk.0.attn_q.weight
n_dimensions = 2
shape = [4096, 4096]
ggml_type = GGML_TYPE_Q2_K
offset = 150544384
==================================================
```
*I used `llama-v1-7b-q2k.gguf` which is generated by @[ikawrakow](https://huggingface.co/ikawrakow)