Running convert fails with BadZipFile (Bad CRC-32) #4365

itsJoKr · 2023-12-07T16:55:30Z

Prerequisites

Please answer the following questions for yourself before submitting an issue.

I am running the latest code. Development is very rapid so there are no tagged versions as of now.
I carefully followed the README.md.
I searched using keywords relevant to my issue to make sure that I am creating a new issue that is not already open (or closed).
I reviewed the Discussions, and have a new bug or useful enhancement to share.

Expected Behavior

The convert runs successfully.

Current Behavior

The convert fails with error like

  File "/Library/Frameworks/Python.framework/Versions/3.12/lib/python3.12/zipfile/__init__.py", line 996, in _update_crc
    raise BadZipFile("Bad CRC-32 for file %r" % self.name)

Environment and Context

Please provide detailed information about your computer setup. This is important in case the issue is not reproducible except for under certain specific conditions.

MacBook Pro, M1 Pro, MacOS Sonoma
Python: 3.12.0
Make: GNU Make 3.81

I tried to run for both OpenLlama-3B and Llama-7B-chat, same error.

Steps to Reproduce

Please provide detailed steps for reproducing the issue. We are not sitting in front of your screen, so the more detail the better.

Clone llama and llama.cpp
Make llama.cpp
Download llama 7B-chat model
Run convert

Failure Logs

llama.cpp$ git log | head 
commit 5a7d3125e7c24f223659b7f0b7aa7736986e92c0

llama.cpp$ python3 --version
Python 3.12.0

llama.cpp$ pip list | egrep "torch|numpy|sentencepiece"
numpy              1.26.2
sentencepiece      0.1.99

The text was updated successfully, but these errors were encountered:

skidd-level-100 · 2023-12-29T22:16:59Z

same issue for all of my models, worked before git pull + recomp

LIMPIX31 · 2024-01-09T09:16:13Z

This seems to have been fixed on the last commit.

skidd-level-100 · 2024-01-11T08:39:16Z

This seems to have been fixed on the last commit.

not for me still 'convert.py path/to/model' and 'convert.py path/to/model.bin' fail with the last to lines being:

raise BadZipFile("Bad CRC-32 for file %r" % self.name)

zipfile.BadZipFile: Bad CRC-32 for file 'archive/data/13'

aspanner · 2024-02-17T01:59:05Z

I've got the same issue. Here's my test run after downloading the model from hf:

python3.12 llama.cpp/convert.py Loading model file aiopsmodel-hfv2/pyt Loading model file aiopsmodel-hfv2/pyt Loading model file aiopsmodel-hfv2/pyt params = Params(n_vocab=32001, Found vocab files: {'tokenizer.model': Loading vocab file 'aiopsmodel-hfv2/to Vocab info: <SentencePieceVocab Special vocab info: <SpecialVocab Permuting layer 0
Permuting layer 1
Permuting layer 2
Permuting layer 3
Permuting layer 4
Permuting layer 5
Permuting layer 6
Permuting layer 7
Permuting layer 8
Permuting layer 9
Permuting layer 10
Permuting layer 11
Permuting layer 12
Permuting layer 13
Permuting layer 14
Permuting layer 15
Permuting layer 16
Permuting layer 17
Permuting layer 18
Permuting layer 19
Permuting layer 20
Permuting layer 21
Permuting layer 22
Permuting layer 23
Permuting layer 24
Permuting layer 25
Permuting layer 26
Permuting layer 27
Permuting layer 28
Permuting layer 29
Permuting layer 30
Permuting layer 31
model.embed_tokens.weight model.layers.0.self_attn.q_proj.weight model.layers.0.self_attn.k_proj.weight model.layers.0.self_attn.v_proj.weight model.layers.0.self_attn.o_proj.weight skipping tensor blk.0.attn_rot_embd
model.layers.0.mlp.gate_proj.weight model.layers.0.mlp.up_proj.weight model.layers.0.mlp.down_proj.weight model.layers.0.input_layernorm.weight model.layers.0.post_attention_layernorm.weight model.layers.1.self_attn.q_proj.weight model.layers.1.self_attn.k_proj.weight model.layers.1.self_attn.v_proj.weight model.layers.1.self_attn.o_proj.weight skipping tensor blk.1.attn_rot_embd
model.layers.1.mlp.gate_proj.weight model.layers.1.mlp.up_proj.weight model.layers.1.mlp.down_proj.weight model.layers.1.input_layernorm.weight model.layers.1.post_attention_layernorm.weight model.layers.2.self_attn.q_proj.weight model.layers.2.self_attn.k_proj.weight model.layers.2.self_attn.v_proj.weight model.layers.2.self_attn.o_proj.weight skipping tensor blk.2.attn_rot_embd
model.layers.2.mlp.gate_proj.weight model.layers.2.mlp.up_proj.weight model.layers.2.mlp.down_proj.weight model.layers.2.input_layernorm.weight model.layers.2.post_attention_layernorm.weight model.layers.3.self_attn.q_proj.weight model.layers.3.self_attn.k_proj.weight model.layers.3.self_attn.v_proj.weight model.layers.3.self_attn.o_proj.weight skipping tensor blk.3.attn_rot_embd
model.layers.3.mlp.gate_proj.weight model.layers.3.mlp.up_proj.weight model.layers.3.mlp.down_proj.weight model.layers.3.input_layernorm.weight model.layers.3.post_attention_layernorm.weight model.layers.4.self_attn.q_proj.weight model.layers.4.self_attn.k_proj.weight model.layers.4.self_attn.v_proj.weight model.layers.4.self_attn.o_proj.weight skipping tensor blk.4.attn_rot_embd
model.layers.4.mlp.gate_proj.weight model.layers.4.mlp.up_proj.weight model.layers.4.mlp.down_proj.weight model.layers.4.input_layernorm.weight model.layers.4.post_attention_layernorm.weight model.layers.5.self_attn.q_proj.weight model.layers.5.self_attn.k_proj.weight model.layers.5.self_attn.v_proj.weight model.layers.5.self_attn.o_proj.weight skipping tensor blk.5.attn_rot_embd
model.layers.5.mlp.gate_proj.weight model.layers.5.mlp.up_proj.weight model.layers.5.mlp.down_proj.weight model.layers.5.input_layernorm.weight model.layers.5.post_attention_layernorm.weight model.layers.6.self_attn.q_proj.weight model.layers.6.self_attn.k_proj.weight model.layers.6.self_attn.v_proj.weight model.layers.6.self_attn.o_proj.weight skipping tensor blk.6.attn_rot_embd
model.layers.6.mlp.gate_proj.weight model.layers.6.mlp.up_proj.weight model.layers.6.mlp.down_proj.weight model.layers.6.input_layernorm.weight model.layers.6.post_attention_layernorm.weight model.layers.7.self_attn.q_proj.weight model.layers.7.self_attn.k_proj.weight model.layers.7.self_attn.v_proj.weight model.layers.7.self_attn.o_proj.weight skipping tensor blk.7.attn_rot_embd
model.layers.7.mlp.gate_proj.weight model.layers.7.mlp.up_proj.weight model.layers.7.mlp.down_proj.weight model.layers.7.input_layernorm.weight model.layers.7.post_attention_layernorm.weight model.layers.8.self_attn.q_proj.weight model.layers.8.self_attn.k_proj.weight model.layers.8.self_attn.v_proj.weight model.layers.8.self_attn.o_proj.weight skipping tensor blk.8.attn_rot_embd
model.layers.8.mlp.gate_proj.weight model.layers.8.mlp.up_proj.weight model.layers.8.mlp.down_proj.weight model.layers.8.input_layernorm.weight model.layers.8.post_attention_layernorm.weight model.layers.9.self_attn.q_proj.weight model.layers.9.self_attn.k_proj.weight model.layers.9.self_attn.v_proj.weight model.layers.9.self_attn.o_proj.weight skipping tensor blk.9.attn_rot_embd
model.layers.9.mlp.gate_proj.weight model.layers.9.mlp.up_proj.weight model.layers.9.mlp.down_proj.weight model.layers.9.input_layernorm.weight model.layers.9.post_attention_layernorm.weight model.layers.10.self_attn.q_proj.weight model.layers.10.self_attn.k_proj.weight model.layers.10.self_attn.v_proj.weight model.layers.10.self_attn.o_proj.weight skipping tensor blk.10.attn_rot_embd File "/home/user1/git/aiops/model_oper main()
File "/home/user1/git/aiops/model_oper OutputFile.write_all(outfile, File "/home/user1/git/aiops/model_oper for i, ((name, lazy_tensor), File "/home/user1/git/aiops/model_oper futures.append(executor.submit(func, ^^^^^^^^^^^^^^
File "/home/user1/git/aiops/model_oper result = futures.pop(0).result()
^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/lib64/python3.12/concurrent return self.__get_result()
^^^^^^^^^^^^^^^^^^^
File "/usr/lib64/python3.12/concurrent raise self._exception
File "/usr/lib64/python3.12/concurrent result = self.fn(*self.args, ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/user1/git/aiops/model_oper tensor = lazy_tensor.load().to_ggml()< ^^^^^^^^^^^^^^^^^^
File "/home/user1/git/aiops/model_oper ret = self._load()
^^^^^^^^^^^^
File "/home/user1/git/aiops/model_oper return self.load().astype(data_type) File "/home/user1/git/aiops/model_oper ret = self._load()
^^^^^^^^^^^^
File "/home/user1/git/aiops/model_oper return lazy_tensor.load().permute(n_head, ^^^^^^^^^^^^^^^^^^
File "/home/user1/git/aiops/model_oper ret = self._load()
^^^^^^^^^^^^
File "/home/user1/git/aiops/model_oper return UnquantizedTensor(storage.load( ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/home/user1/git/aiops/model_oper data = fp.read(size)
^^^^^^^^^^^^^
File "/usr/lib64/python3.12/zipfile/ ^^^^^^^^^^^^^^
File "/usr/lib64/python3.12/zipfile/ File "/usr/lib64/python3.12/zipfile/ orch_model-00001-of-00002.bin
orch_model-00001-of-00002.bin
orch_model-00002-of-00002.bin
n_embd=4096, n_layer=32, n_ctx=4096, n_ff=11008, n_head=32, n_head_kv=32, n_experts=None, n_experts_used=None, f_norm_eps=1e-05, rope_scaling_type=None, f_rope_freq_base=None, f_rope_scale=None, n_orig_ctx=None, rope_finetuned=None, ftype=<GGMLFileType.MostlyQ8_0: 7>, path_model=PosixPath('aiopsmodel-hfv2'))
PosixPath('aiopsmodel-hfv2/tokenizer.model'), 'vocab.json': None, 'tokenizer.json': PosixPath('aiopsmodel-hfv2/tokenizer.json')}
kenizer.model', type 'spm'
with 32000 base tokens and 1 added tokens>
with 0 merges, special tokens {'bos': 1, 'eos': 2, 'unk': 0, 'pad': 0}, add special tokens unset>
-> token_embd.weight | F16 | [32000, 4096]
-> blk.0.attn_q.weight | F16 | [4096, 4096]
-> blk.0.attn_k.weight | F16 | [4096, 4096]
-> blk.0.attn_v.weight | F16 | [4096, 4096]
-> blk.0.attn_output.weight | F16 | [4096, 4096]
-> blk.0.ffn_gate.weight | F16 | [11008, 4096]
-> blk.0.ffn_up.weight | F16 | [11008, 4096]
-> blk.0.ffn_down.weight | F16 | [4096, 11008]
-> blk.0.attn_norm.weight | F16 | [4096]
-> blk.0.ffn_norm.weight | F16 | [4096]
-> blk.1.attn_q.weight | F16 | [4096, 4096]
-> blk.1.attn_k.weight | F16 | [4096, 4096]
-> blk.1.attn_v.weight | F16 | [4096, 4096]
-> blk.1.attn_output.weight | F16 | [4096, 4096]
-> blk.1.ffn_gate.weight | F16 | [11008, 4096]
-> blk.1.ffn_up.weight | F16 | [11008, 4096]
-> blk.1.ffn_down.weight | F16 | [4096, 11008]
-> blk.1.attn_norm.weight | F16 | [4096]
-> blk.1.ffn_norm.weight | F16 | [4096]
-> blk.2.attn_q.weight | F16 | [4096, 4096]
-> blk.2.attn_k.weight | F16 | [4096, 4096]
-> blk.2.attn_v.weight | F16 | [4096, 4096]
-> blk.2.attn_output.weight | F16 | [4096, 4096]
-> blk.2.ffn_gate.weight | F16 | [11008, 4096]
-> blk.2.ffn_up.weight | F16 | [11008, 4096]
-> blk.2.ffn_down.weight | F16 | [4096, 11008]
-> blk.2.attn_norm.weight | F16 | [4096]
-> blk.2.ffn_norm.weight | F16 | [4096]
-> blk.3.attn_q.weight | F16 | [4096, 4096]
-> blk.3.attn_k.weight | F16 | [4096, 4096]
-> blk.3.attn_v.weight | F16 | [4096, 4096]
-> blk.3.attn_output.weight | F16 | [4096, 4096]
-> blk.3.ffn_gate.weight | F16 | [11008, 4096]
-> blk.3.ffn_up.weight | F16 | [11008, 4096]
-> blk.3.ffn_down.weight | F16 | [4096, 11008]
-> blk.3.attn_norm.weight | F16 | [4096]
-> blk.3.ffn_norm.weight | F16 | [4096]
-> blk.4.attn_q.weight | F16 | [4096, 4096]
-> blk.4.attn_k.weight | F16 | [4096, 4096]
-> blk.4.attn_v.weight | F16 | [4096, 4096]
-> blk.4.attn_output.weight | F16 | [4096, 4096]
-> blk.4.ffn_gate.weight | F16 | [11008, 4096]
-> blk.4.ffn_up.weight | F16 | [11008, 4096]
-> blk.4.ffn_down.weight | F16 | [4096, 11008]
-> blk.4.attn_norm.weight | F16 | [4096]
-> blk.4.ffn_norm.weight | F16 | [4096]
-> blk.5.attn_q.weight | F16 | [4096, 4096]
-> blk.5.attn_k.weight | F16 | [4096, 4096]
-> blk.5.attn_v.weight | F16 | [4096, 4096]
-> blk.5.attn_output.weight | F16 | [4096, 4096]
-> blk.5.ffn_gate.weight | F16 | [11008, 4096]
-> blk.5.ffn_up.weight | F16 | [11008, 4096]
-> blk.5.ffn_down.weight | F16 | [4096, 11008]
-> blk.5.attn_norm.weight | F16 | [4096]
-> blk.5.ffn_norm.weight | F16 | [4096]
-> blk.6.attn_q.weight | F16 | [4096, 4096]
-> blk.6.attn_k.weight | F16 | [4096, 4096]
-> blk.6.attn_v.weight | F16 | [4096, 4096]
-> blk.6.attn_output.weight | F16 | [4096, 4096]
-> blk.6.ffn_gate.weight | F16 | [11008, 4096]
-> blk.6.ffn_up.weight | F16 | [11008, 4096]
-> blk.6.ffn_down.weight | F16 | [4096, 11008]
-> blk.6.attn_norm.weight | F16 | [4096]
-> blk.6.ffn_norm.weight | F16 | [4096]
-> blk.7.attn_q.weight | F16 | [4096, 4096]
-> blk.7.attn_k.weight | F16 | [4096, 4096]
-> blk.7.attn_v.weight | F16 | [4096, 4096]
-> blk.7.attn_output.weight | F16 | [4096, 4096]
-> blk.7.ffn_gate.weight | F16 | [11008, 4096]
-> blk.7.ffn_up.weight | F16 | [11008, 4096]
-> blk.7.ffn_down.weight | F16 | [4096, 11008]
-> blk.7.attn_norm.weight | F16 | [4096]
-> blk.7.ffn_norm.weight | F16 | [4096]
-> blk.8.attn_q.weight | F16 | [4096, 4096]
-> blk.8.attn_k.weight | F16 | [4096, 4096]
-> blk.8.attn_v.weight | F16 | [4096, 4096]
-> blk.8.attn_output.weight | F16 | [4096, 4096]
-> blk.8.ffn_gate.weight | F16 | [11008, 4096]
-> blk.8.ffn_up.weight | F16 | [11008, 4096]
-> blk.8.ffn_down.weight | F16 | [4096, 11008]
-> blk.8.attn_norm.weight | F16 | [4096]
-> blk.8.ffn_norm.weight | F16 | [4096]
-> blk.9.attn_q.weight | F16 | [4096, 4096]
-> blk.9.attn_k.weight | F16 | [4096, 4096]
-> blk.9.attn_v.weight | F16 | [4096, 4096]
-> blk.9.attn_output.weight | F16 | [4096, 4096]
-> blk.9.ffn_gate.weight | F16 | [11008, 4096]
-> blk.9.ffn_up.weight | F16 | [11008, 4096]
-> blk.9.ffn_down.weight | F16 | [4096, 11008]
-> blk.9.attn_norm.weight | F16 | [4096]
-> blk.9.ffn_norm.weight | F16 | [4096]
-> blk.10.attn_q.weight | F16 | [4096, 4096]
-> blk.10.attn_k.weight | F16 | [4096, 4096]
-> blk.10.attn_v.weight | F16 | [4096, 4096]
-> blk.10.attn_output.weight | F16 | [4096, 4096]
r> -> blk.10.ffn_gate.weight | F16 | [11008, 4096]
-> blk.10.ffn_up.weight | F16 | [11008, 4096]
-> blk.10.ffn_down.weight | F16 | [4096, 11008]
-> blk.10.attn_norm.weight | F16 | [4096]
-> blk.10.ffn_norm.weight | F16 | [4096]
-> blk.11.attn_q.weight | F16 | [4096, 4096]
-> blk.11.attn_k.weight | F16 | [4096, 4096]
-> blk.11.attn_v.weight | F16 | [4096, 4096]
-> blk.11.attn_output.weight | F16 | [4096, 4096]
r> -> blk.11.ffn_gate.weight | F16 | [11008, 4096]
-> blk.11.ffn_up.weight | F16 | [11008, 4096]
-> blk.11.ffn_down.weight | F16 | [4096, 11008]
-> blk.11.attn_norm.weight | F16 | [4096]
-> blk.11.ffn_norm.weight | F16 | [4096]
-> blk.12.attn_q.weight | F16 | [4096, 4096]
-> blk.12.attn_k.weight | F16 | [4096, 4096]
-> blk.12.attn_v.weight | F16 | [4096, 4096]
-> blk.12.attn_output.weight | F16 | [4096, 4096]
r> -> blk.12.ffn_gate.weight | F16 | [11008, 4096]
-> blk.12.ffn_up.weight | F16 | [11008, 4096]
-> blk.12.ffn_down.weight | F16 | [4096, 11008]
-> blk.12.attn_norm.weight | F16 | [4096]
-> blk.12.ffn_norm.weight | F16 | [4096]
-> blk.13.attn_q.weight | F16 | [4096, 4096]
-> blk.13.attn_k.weight | F16 | [4096, 4096]
-> blk.13.attn_v.weight | F16 | [4096, 4096]
-> blk.13.attn_output.weight | F16 | [4096, 4096]
r> -> blk.13.ffn_gate.weight | F16 | [11008, 4096]
-> blk.13.ffn_up.weight | F16 | [11008, 4096]
-> blk.13.ffn_down.weight | F16 | [4096, 11008]
-> blk.13.attn_norm.weight | F16 | [4096]
-> blk.13.ffn_norm.weight | F16 | [4096]
-> blk.14.attn_q.weight | F16 | [4096, 4096]
-> blk.14.attn_k.weight | F16 | [4096, 4096]
-> blk.14.attn_v.weight | F16 | [4096, 4096]
-> blk.14.attn_output.weight | F16 | [4096, 4096]
r> -> blk.14.ffn_gate.weight | F16 | [11008, 4096]
-> blk.14.ffn_up.weight | F16 | [11008, 4096]
-> blk.14.ffn_down.weight | F16 | [4096, 11008]
-> blk.14.attn_norm.weight | F16 | [4096]
-> blk.14.ffn_norm.weight | F16 | [4096]
-> blk.15.attn_q.weight | F16 | [4096, 4096]
-> blk.15.attn_k.weight | F16 | [4096, 4096]
-> blk.15.attn_v.weight | F16 | [4096, 4096]
-> blk.15.attn_output.weight | F16 | [4096, 4096]
r> -> blk.15.ffn_gate.weight | F16 | [11008, 4096]
-> blk.15.ffn_up.weight | F16 | [11008, 4096]
-> blk.15.ffn_down.weight | F16 | [4096, 11008]
-> blk.15.attn_norm.weight | F16 | [4096]
-> blk.15.ffn_norm.weight | F16 | [4096]
-> blk.16.attn_q.weight | F16 | [4096, 4096]
-> blk.16.attn_k.weight | F16 | [4096, 4096]
-> blk.16.attn_v.weight | F16 | [4096, 4096]
-> blk.16.attn_output.weight | F16 | [4096, 4096]
r> -> blk.16.ffn_gate.weight | F16 | [11008, 4096]
-> blk.16.ffn_up.weight | F16 | [11008, 4096]
-> blk.16.ffn_down.weight | F16 | [4096, 11008]
-> blk.16.attn_norm.weight | F16 | [4096]
-> blk.16.ffn_norm.weight | F16 | [4096]
-> blk.17.attn_q.weight | F16 | [4096, 4096]
-> blk.17.attn_k.weight | F16 | [4096, 4096]
-> blk.17.attn_v.weight | F16 | [4096, 4096]
-> blk.17.attn_output.weight | F16 | [4096, 4096]
r> -> blk.17.ffn_gate.weight | F16 | [11008, 4096]
-> blk.17.ffn_up.weight | F16 | [11008, 4096]
-> blk.17.ffn_down.weight | F16 | [4096, 11008]
-> blk.17.attn_norm.weight | F16 | [4096]
-> blk.17.ffn_norm.weight | F16 | [4096]
-> blk.18.attn_q.weight | F16 | [4096, 4096]
-> blk.18.attn_k.weight | F16 | [4096, 4096]
-> blk.18.attn_v.weight | F16 | [4096, 4096]
-> blk.18.attn_output.weight | F16 | [4096, 4096]
r> -> blk.18.ffn_gate.weight | F16 | [11008, 4096]
-> blk.18.ffn_up.weight | F16 | [11008, 4096]
-> blk.18.ffn_down.weight | F16 | [4096, 11008]
-> blk.18.attn_norm.weight | F16 | [4096]
-> blk.18.ffn_norm.weight | F16 | [4096]
-> blk.19.attn_q.weight | F16 | [4096, 4096]
-> blk.19.attn_k.weight | F16 | [4096, 4096]
-> blk.19.attn_v.weight | F16 | [4096, 4096]
-> blk.19.attn_output.weight | F16 | [4096, 4096]
r> -> blk.19.ffn_gate.weight | F16 | [11008, 4096]
-> blk.19.ffn_up.weight | F16 | [11008, 4096]
-> blk.19.ffn_down.weight | F16 | [4096, 11008]
-> blk.19.attn_norm.weight | F16 | [4096]
-> blk.19.ffn_norm.weight | F16 | [4096]
-> blk.20.attn_q.weight | F16 | [4096, 4096]
-> blk.20.attn_k.weight | F16 | [4096, 4096]
-> blk.20.attn_v.weight | F16 | [4096, 4096]
-> blk.20.attn_output.weight | F16 | [4096, 4096]
r> -> blk.20.ffn_gate.weight | F16 | [11008, 4096]
-> blk.20.ffn_up.weight | F16 | [11008, 4096]
-> blk.20.ffn_down.weight | F16 | [4096, 11008]
-> blk.20.attn_norm.weight | F16 | [4096]
-> blk.20.ffn_norm.weight | F16 | [4096]
-> blk.21.attn_q.weight | F16 | [4096, 4096]
-> blk.21.attn_k.weight | F16 | [4096, 4096]
-> blk.21.attn_v.weight | F16 | [4096, 4096]
-> blk.21.attn_output.weight | F16 | [4096, 4096]
r> -> blk.21.ffn_gate.weight | F16 | [11008, 4096]
-> blk.21.ffn_up.weight | F16 | [11008, 4096]
-> blk.21.ffn_down.weight | F16 | [4096, 11008]
-> blk.21.attn_norm.weight | F16 | [4096]
-> blk.21.ffn_norm.weight | F16 | [4096]
-> blk.22.attn_q.weight | F16 | [4096, 4096]
-> blk.22.attn_k.weight | F16 | [4096, 4096]
-> blk.22.attn_v.weight | F16 | [4096, 4096]
-> blk.22.attn_output.weight | F16 | [4096, 4096]
r> -> blk.22.ffn_gate.weight | F16 | [11008, 4096]
-> blk.22.ffn_up.weight | F16 | [11008, 4096]
-> blk.22.ffn_down.weight | F16 | [4096, 11008]
-> blk.22.attn_norm.weight | F16 | [4096]
-> blk.22.ffn_norm.weight | F16 | [4096]
-> blk.23.attn_q.weight | F16 | [4096, 4096]
-> blk.23.attn_k.weight | F16 | [4096, 4096]
-> blk.23.attn_v.weight | F16 | [4096, 4096]
-> blk.23.attn_output.weight | F16 | [4096, 4096]
r> -> blk.23.ffn_gate.weight | F16 | [11008, 4096]
-> blk.23.ffn_up.weight | F16 | [11008, 4096]
-> blk.23.ffn_down.weight | F16 | [4096, 11008]
-> blk.23.attn_norm.weight | F16 | [4096]
-> blk.23.ffn_norm.weight | F16 | [4096]
-> blk.24.attn_q.weight | F16 | [4096, 4096]
-> blk.24.attn_k.weight | F16 | [4096, 4096]
-> blk.24.attn_v.weight | F16 | [4096, 4096]
-> blk.24.attn_output.weight | F16 | [4096, 4096]
r> -> blk.24.ffn_gate.weight | F16 | [11008, 4096]
-> blk.24.ffn_up.weight | F16 | [11008, 4096]
-> blk.24.ffn_down.weight | F16 | [4096, 11008]
-> blk.24.attn_norm.weight | F16 | [4096]
-> blk.24.ffn_norm.weight | F16 | [4096]
-> blk.25.attn_q.weight | F16 | [4096, 4096]
-> blk.25.attn_k.weight | F16 | [4096, 4096]
-> blk.25.attn_v.weight | F16 | [4096, 4096]
-> blk.25.attn_output.weight | F16 | [4096, 4096]
r> -> blk.25.ffn_gate.weight | F16 | [11008, 4096]
-> blk.25.ffn_up.weight | F16 | [11008, 4096]
-> blk.25.ffn_down.weight | F16 | [4096, 11008]
-> blk.25.attn_norm.weight | F16 | [4096]
-> blk.25.ffn_norm.weight | F16 | [4096]
-> blk.26.attn_q.weight | F16 | [4096, 4096]
-> blk.26.attn_k.weight | F16 | [4096, 4096]
-> blk.26.attn_v.weight | F16 | [4096, 4096]
-> blk.26.attn_output.weight | F16 | [4096, 4096]
r> -> blk.26.ffn_gate.weight | F16 | [11008, 4096]
-> blk.26.ffn_up.weight | F16 | [11008, 4096]
-> blk.26.ffn_down.weight | F16 | [4096, 11008]
-> blk.26.attn_norm.weight | F16 | [4096]
-> blk.26.ffn_norm.weight | F16 | [4096]
-> blk.27.attn_q.weight | F16 | [4096, 4096]
-> blk.27.attn_k.weight | F16 | [4096, 4096]
-> blk.27.attn_v.weight | F16 | [4096, 4096]
-> blk.27.attn_output.weight | F16 | [4096, 4096]
r> -> blk.27.ffn_gate.weight | F16 | [11008, 4096]
-> blk.27.ffn_up.weight | F16 | [11008, 4096]
-> blk.27.ffn_down.weight | F16 | [4096, 11008]
-> blk.27.attn_norm.weight | F16 | [4096]
-> blk.27.ffn_norm.weight | F16 | [4096]
-> blk.28.attn_q.weight | F16 | [4096, 4096]
-> blk.28.attn_k.weight | F16 | [4096, 4096]
-> blk.28.attn_v.weight | F16 | [4096, 4096]
-> blk.28.attn_output.weight | F16 | [4096, 4096]
r> -> blk.28.ffn_gate.weight | F16 | [11008, 4096]
-> blk.28.ffn_up.weight | F16 | [11008, 4096]
-> blk.28.ffn_down.weight | F16 | [4096, 11008]
-> blk.28.attn_norm.weight | F16 | [4096]
-> blk.28.ffn_norm.weight | F16 | [4096]
-> blk.29.attn_q.weight | F16 | [4096, 4096]
-> blk.29.attn_k.weight | F16 | [4096, 4096]
-> blk.29.attn_v.weight | F16 | [4096, 4096]
-> blk.29.attn_output.weight | F16 | [4096, 4096]
r> -> blk.29.ffn_gate.weight | F16 | [11008, 4096]
-> blk.29.ffn_up.weight | F16 | [11008, 4096]
-> blk.29.ffn_down.weight | F16 | [4096, 11008]
-> blk.29.attn_norm.weight | F16 | [4096]
-> blk.29.ffn_norm.weight | F16 | [4096]
-> blk.30.attn_q.weight | F16 | [4096, 4096]
-> blk.30.attn_k.weight | F16 | [4096, 4096]
-> blk.30.attn_v.weight | F16 | [4096, 4096]
-> blk.30.attn_output.weight | F16 | [4096, 4096]
r> -> blk.30.ffn_gate.weight | F16 | [11008, 4096]
-> blk.30.ffn_up.weight | F16 | [11008, 4096]
-> blk.30.ffn_down.weight | F16 | [4096, 11008]
-> blk.30.attn_norm.weight | F16 | [4096]
-> blk.30.ffn_norm.weight | F16 | [4096]
-> blk.31.attn_q.weight | F16 | [4096, 4096]
-> blk.31.attn_k.weight | F16 | [4096, 4096]
-> blk.31.attn_v.weight | F16 | [4096, 4096]
-> blk.31.attn_output.weight | F16 | [4096, 4096]
r> -> blk.31.ffn_gate.weight | F16 | [11008, 4096]
-> blk.31.ffn_up.weight | F16 | [11008, 4096]
-> blk.31.ffn_down.weight | F16 | [4096, 11008]
-> blk.31.attn_norm.weight | F16 | [4096]
-> blk.31.ffn_norm.weight | F16 | [4096]
-> output_norm.weight | F16 | [4096]
-> output.weight | F16 | [32000, 4096]
format 7
model matches vocab size without it.
Endian only
bos to 1
eos to 2
unk to 0
pad to 0
ations/llama.cpp/convert.py", line 1483, in
ations/llama.cpp/convert.py", line 1477, in main
ftype, params, model, vocab, special_vocab,
ations/llama.cpp/convert.py", line 1144, in write_all
ndarray) in enumerate(zip(model.items(), ndarrays)):
ations/llama.cpp/convert.py", line 918, in bounded_parallel_map
next(iterable)))
ations/llama.cpp/convert.py", line 924, in bounded_parallel_map
/futures/_base.py", line 449, in result
/futures/_base.py", line 401, in __get_result
/futures/thread.py", line 58, in run
**self.kwargs)
ations/llama.cpp/convert.py", line 1101, in do_item
br> ations/llama.cpp/convert.py", line 649, in load
ations/llama.cpp/convert.py", line 659, in load
r> ations/llama.cpp/convert.py", line 649, in load
ations/llama.cpp/convert.py", line 737, in load
n_head_kv)
ations/llama.cpp/convert.py", line 649, in load
ations/llama.cpp/convert.py", line 809, in load
storage_offset, elm_count).reshape(size))
^
ations/llama.cpp/convert.py", line 796, in load
trong>init.py", line 979, in read
trong>init.py", line 1069, in _read1
trong>init.py", line 997, in _update_crc
for file %r" % self.name)
for file 'pytorch_model-00001-of-00002/data/2'

It did start with the requirments.txt not installing. I had to replace the ~= with >= to make it install.
I am using a clean virtual python environement.

JaCraig · 2024-02-22T15:32:25Z

It did start with the requirments.txt not installing. I had to replace the ~= with >= to make it install. I am using a clean virtual python environement.

This is due to ImpImporter being removed recently from pip (the Python thread where ImpImporter is deprecated). The zipfile issue also seems to be 3.12 specific. Switching to 3.10 in my environment fixed the issue for me.

wdavo · 2024-03-04T10:46:49Z

Thanks @JaCraig - Can confirm that switching from 3.12 to 3.10 fixed the issue for me as well.

harshil21 · 2024-03-09T00:47:14Z

Facing the same issue. I don't want to downgrade my python just for this. I think llama.cpp should support Python 3.12. The latest pytorch version (2.2.1) now also supports Python 3.12 (pytorch/pytorch#110436 (comment)) so it shouldn't be a problem to support it.

MarcoLv412 · 2024-03-25T22:21:14Z

Same issue here, on windows 10 PC:
Traceback (most recent call last):
File "D:\eosusa\cpp\llama.cpp\convert.py", line 1486, in
main()
File "D:\eosusa\cpp\llama.cpp\convert.py", line 1480, in main
OutputFile.write_all(outfile, ftype, params, model, vocab, special_vocab,
File "D:\eosusa\cpp\llama.cpp\convert.py", line 1162, in write_all
of.write_tensor_data(ftype, model, concurrency)
File "D:\eosusa\cpp\llama.cpp\convert.py", line 1093, in write_tensor_data
for i, ((name, lazy_tensor), ndarray) in enumerate(zip(model.items(), ndarrays)):
File "D:\eosusa\cpp\llama.cpp\convert.py", line 941, in bounded_parallel_map
result = futures.pop(0).result()
^^^^^^^^^^^^^^^^^^^^^^^
File "C:\ProgramData\miniconda3\Lib\concurrent\futures_base.py", line 449, in result
return self.__get_result()
^^^^^^^^^^^^^^^^^^^
File "C:\ProgramData\miniconda3\Lib\concurrent\futures_base.py", line 401, in __get_result
raise self._exception
File "C:\ProgramData\miniconda3\Lib\concurrent\futures\thread.py", line 58, in run
result = self.fn(*self.args, **self.kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "D:\eosusa\cpp\llama.cpp\convert.py", line 1126, in do_item
tensor = lazy_tensor.load().to_ggml()
^^^^^^^^^^^^^^^^^^
File "D:\eosusa\cpp\llama.cpp\convert.py", line 666, in load
ret = self._load()
^^^^^^^^^^^^
File "D:\eosusa\cpp\llama.cpp\convert.py", line 676, in load
return self.load().astype(data_type)
^^^^^^^^^^^
File "D:\eosusa\cpp\llama.cpp\convert.py", line 666, in load
ret = self.load()
^^^^^^^^^^^^
File "D:\eosusa\cpp\llama.cpp\convert.py", line 826, in load
return UnquantizedTensor(storage.load(storage_offset, elm_count).reshape(size))
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "D:\eosusa\cpp\llama.cpp\convert.py", line 810, in load
fp = self.zip_file.open(info)
^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\ProgramData\miniconda3\Lib\zipfile_init.py", line 1643, in open
raise BadZipFile(f"Overlapped entries: {zinfo.orig_filename!r} (possible zip bomb)")
zipfile.BadZipFile: Overlapped entries: 'consolidated.00/data/73' (possible zip bomb)

python --version
Python 3.12.2

All requriements.txt installed.

ifoxhz · 2024-04-28T01:57:34Z

the same issue on macos:
data = self._read1(n)
File "/usr/local/Cellar/[email protected]/3.9.13_2/Frameworks/Python.framework/Versions/3.9/lib/python3.9/zipfile.py", line 1014, in _read1
self._update_crc(data)
File "/usr/local/Cellar/[email protected]/3.9.13_2/Frameworks/Python.framework/Versions/3.9/lib/python3.9/zipfile.py", line 942, in _update_crc
raise BadZipFile("Bad CRC-32 for file %r" % self.name)
zipfile.BadZipFile: Bad CRC-32 for file 'pytorch_model-00001-of-00002/data/144'

llmama.cpp version:
commit 4e96a81 (HEAD -> master, origin/master, origin/HEAD)
Author: Anas Ahouzi [email protected]
Date: Tue Apr 23 02:53:18 2024 +0200

python3.10

Leonidas-Li · 2024-05-08T09:27:35Z

It did start with the requirments.txt not installing. I had to replace the ~= with >= to make it install. I am using a clean virtual python environement.

This is due to ImpImporter being removed recently from pip (the Python thread where ImpImporter is deprecated). The zipfile issue also seems to be 3.12 specific. Switching to 3.10 in my environment fixed the issue for me.

Thanks @JaCraig Solved my problem.

QingtaoLi1 · 2024-05-20T07:10:09Z

Same issue here, on windows 10 PC: Traceback (most recent call last): File "D:\eosusa\cpp\llama.cpp\convert.py", line 1486, in main() File "D:\eosusa\cpp\llama.cpp\convert.py", line 1480, in main OutputFile.write_all(outfile, ftype, params, model, vocab, special_vocab, File "D:\eosusa\cpp\llama.cpp\convert.py", line 1162, in write_all of.write_tensor_data(ftype, model, concurrency) File "D:\eosusa\cpp\llama.cpp\convert.py", line 1093, in write_tensor_data for i, ((name, lazy_tensor), ndarray) in enumerate(zip(model.items(), ndarrays)): File "D:\eosusa\cpp\llama.cpp\convert.py", line 941, in bounded_parallel_map result = futures.pop(0).result() ^^^^^^^^^^^^^^^^^^^^^^^ File "C:\ProgramData\miniconda3\Lib\concurrent\futures_base.py", line 449, in result return self.__get_result() ^^^^^^^^^^^^^^^^^^^ File "C:\ProgramData\miniconda3\Lib\concurrent\futures_base.py", line 401, in __get_result raise self._exception File "C:\ProgramData\miniconda3\Lib\concurrent\futures\thread.py", line 58, in run result = self.fn(*self.args, **self.kwargs) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "D:\eosusa\cpp\llama.cpp\convert.py", line 1126, in do_item tensor = lazy_tensor.load().to_ggml() ^^^^^^^^^^^^^^^^^^ File "D:\eosusa\cpp\llama.cpp\convert.py", line 666, in load ret = self.load() ^^^^^^^^^^^^ File "D:\eosusa\cpp\llama.cpp\convert.py", line 676, in load return self.load().astype(data_type) ^^^^^^^^^^^ File "D:\eosusa\cpp\llama.cpp\convert.py", line 666, in load ret = self.load() ^^^^^^^^^^^^ File "D:\eosusa\cpp\llama.cpp\convert.py", line 826, in load return UnquantizedTensor(storage.load(storage_offset, elm_count).reshape(size)) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "D:\eosusa\cpp\llama.cpp\convert.py", line 810, in load fp = self.zip_file.open(info) ^^^^^^^^^^^^^^^^^^^^^^^^ File "C:\ProgramData\miniconda3\Lib\zipfile__init.py", line 1643, in open raise BadZipFile(f"Overlapped entries: {zinfo.orig_filename!r} (possible zip bomb)") zipfile.BadZipFile: Overlapped entries: 'consolidated.00/data/73' (possible zip bomb)

python --version Python 3.12.2

All requriements.txt installed.

@MarcoLv412 Same error message here on Windows 11, python=3.8. Try adding "--concurrency=1" which works for me. Seems some problems in multi-thread zipfile reading.

teleprint-me · 2024-05-20T21:16:03Z

@mofosyne This is a bug. It is python specific though. Add to docs for now?

Using safetensors is a valid workaround.

github-actions · 2024-07-06T01:06:42Z

This issue was closed because it has been inactive for 14 days since being marked as stale.

LianaT · 2024-09-01T21:37:48Z

Is this still an issue or has it been fixed because it's marked as closed? I'd rather stick with Python 3.12 which I already have on macos, instead of having to switch to python 3.10

itsJoKr added the bug-unconfirmed label Dec 7, 2023

github-actions bot added the stale label Apr 25, 2024

github-actions bot removed the stale label Apr 29, 2024

github-actions bot added the stale label Jun 21, 2024

github-actions bot closed this as completed Jul 6, 2024

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Running convert fails with BadZipFile (Bad CRC-32) #4365

Running convert fails with BadZipFile (Bad CRC-32) #4365

itsJoKr commented Dec 7, 2023

skidd-level-100 commented Dec 29, 2023

LIMPIX31 commented Jan 9, 2024 •

edited

Loading

skidd-level-100 commented Jan 11, 2024

aspanner commented Feb 17, 2024

JaCraig commented Feb 22, 2024 •

edited

Loading

wdavo commented Mar 4, 2024

harshil21 commented Mar 9, 2024

MarcoLv412 commented Mar 25, 2024

ifoxhz commented Apr 28, 2024 •

edited

Loading

Leonidas-Li commented May 8, 2024

QingtaoLi1 commented May 20, 2024 •

edited

Loading

teleprint-me commented May 20, 2024 •

edited

Loading

github-actions bot commented Jul 6, 2024

LianaT commented Sep 1, 2024

Running convert fails with BadZipFile (Bad CRC-32) #4365

Running convert fails with BadZipFile (Bad CRC-32) #4365

Comments

itsJoKr commented Dec 7, 2023

Prerequisites

Expected Behavior

Current Behavior

Environment and Context

Steps to Reproduce

Failure Logs

skidd-level-100 commented Dec 29, 2023

LIMPIX31 commented Jan 9, 2024 • edited Loading

skidd-level-100 commented Jan 11, 2024

aspanner commented Feb 17, 2024

JaCraig commented Feb 22, 2024 • edited Loading

wdavo commented Mar 4, 2024

harshil21 commented Mar 9, 2024

MarcoLv412 commented Mar 25, 2024

ifoxhz commented Apr 28, 2024 • edited Loading

Leonidas-Li commented May 8, 2024

QingtaoLi1 commented May 20, 2024 • edited Loading

teleprint-me commented May 20, 2024 • edited Loading

github-actions bot commented Jul 6, 2024

LianaT commented Sep 1, 2024

LIMPIX31 commented Jan 9, 2024 •

edited

Loading

JaCraig commented Feb 22, 2024 •

edited

Loading

ifoxhz commented Apr 28, 2024 •

edited

Loading

QingtaoLi1 commented May 20, 2024 •

edited

Loading

teleprint-me commented May 20, 2024 •

edited

Loading