llama
llama copied to clipboard
meta-llama
download.sh didn't work well
Before submitting a bug, please make sure the issue hasn't been already addressed by searching through the FAQs and existing/past issues
Describe the bug
run download.sh directly, it could download USE_POLICY.md、tokenizer.model successfully, but failed to download tokenizer_checklist.chk
Minimal reproducible example
<Remember to wrap the code in ```triple-quotes blocks```>
# sample code to repro the bug
Output
<Remember to wrap the output in ```triple-quotes blocks```>
root@iZj6c6gt4172cj579s3pkvZ:~/llama# ./download.sh
Enter the URL from email: https://download5.llamameta.net/*?Policy=<policy>&Key-Pair-Id=<id>&Download-Request-ID=<request_id>
Enter the list of models to download without spaces (7B,13B,70B,7B-chat,13B-chat,70B-chat), or press Enter for all:
Downloading LICENSE and Acceptable Usage Policy
--2024-04-25 10:35:52-- https://download5.llamameta.net/LICENSE?Policy=<policy>&Key-Pair-Id=<id>&Download-Request-ID=<request_id>
Resolving download5.llamameta.net (download5.llamameta.net)... 13.224.163.120, 13.224.163.51, 13.224.163.33, ...
Connecting to download5.llamameta.net (download5.llamameta.net)|13.224.163.120|:443... connected.
HTTP request sent, awaiting response... 416 Requested Range Not Satisfiable
The file is already fully retrieved; nothing to do.
--2024-04-25 10:35:52-- https://download5.llamameta.net/USE_POLICY.md?Policy=<policy>&Key-Pair-Id=<id>&Download-Request-ID=<request_id>
Resolving download5.llamameta.net (download5.llamameta.net)... 13.224.163.10, 13.224.163.33, 13.224.163.51, ...
Connecting to download5.llamameta.net (download5.llamameta.net)|13.224.163.10|:443... connected.
HTTP request sent, awaiting response... 206 Partial Content
Length: 4790 (4.7K), 24 remaining [text/markdown]
Saving to: ‘./USE_POLICY.md’
./USE_POLICY.md 100%[++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++>] 4.68K --.-KB/s in 0s
2024-04-25 10:35:53 (1.44 MB/s) - ‘./USE_POLICY.md’ saved [4790/4790]
Downloading tokenizer
--2024-04-25 10:35:53-- https://download5.llamameta.net/tokenizer.model?Policy=<policy>&Key-Pair-Id=<id>&Download-Request-ID=<request_id>
Resolving download5.llamameta.net (download5.llamameta.net)... 13.224.163.10, 13.224.163.33, 13.224.163.51, ...
Connecting to download5.llamameta.net (download5.llamameta.net)|13.224.163.10|:443... connected.
HTTP request sent, awaiting response... 200 OK
Length: 499723 (488K) [binary/octet-stream]
Saving to: ‘./tokenizer.model’
./tokenizer.model 100%[==================================================================================================================>] 488.01K 411KB/s in 1.2s
2024-04-25 10:35:55 (411 KB/s) - ‘./tokenizer.model’ saved [499723/499723]
--2024-04-25 10:35:55-- https://download5.llamameta.net/tokenizer_checklist.chk?Policy=<policy>&Key-Pair-Id=<id>&Download-Request-ID=<request_id>
Resolving download5.llamameta.net (download5.llamameta.net)... 13.224.163.10, 13.224.163.33, 13.224.163.51, ...
Connecting to download5.llamameta.net (download5.llamameta.net)|13.224.163.10|:443... connected.
HTTP request sent, awaiting response... 403 Forbidden
2024-04-25 10:35:56 ERROR 403: Forbidden.
Runtime Environment
- Model: llama3
- Using via huggingface?: no
- OS: Linux
- GPU VRAM:
- Number of GPUs:
- GPU Make: [eg: Nvidia, AMD, Intel]
Additional context Add any other context about the problem or environment here.
Have no idea but maybe you could try delete the whole repo and git clone it again. The first time I ran download.sh it just says ERROR 403: Forbidden. like the last line of your error message, after I delete the whole repo and git clone it again it works quite well.
Tried but doesn't work. does it mean we can't download the model in China ?
Tried but doesn't work. does it mean we can't download the model in China ?
Don't think so. I deployed it on remote server in my school and there wasn't VPN or proxy or something like that on the device, but it worked.
hey bro,have you solved the problem now? i have the same problem.
use ollama instead