CloverLM-8b โ training checkpoints
Raw PyTorch training checkpoints for CloverLM-8b (climbmix run). Each training step lives on its own git branch so you can download exactly the checkpoint you need without pulling the whole history.
| Training step | Branch | File |
|---|---|---|
| 20000 | step-20000 |
climbmix-8b-20000.pt |
| 40000 | step-40000 |
climbmix-8b-40000.pt |
| 60000 | step-60000 |
climbmix-8b-60000.pt |
| 80000 | step-80000 |
climbmix-8b-80000.pt |
| 100000 | step-100000 |
climbmix-8b-100000.pt |
| 120000 | step-120000 |
climbmix-8b-120000.pt |
| 140000 | step-140000 |
climbmix-8b-140000.pt |
| 160000 | step-160000 |
climbmix-8b-160000.pt |
| 165000 | step-165000 |
climbmix-8b-165000.pt |
| 180000 | step-180000 |
climbmix-8b-180000.pt |
| 200000 | step-200000 |
climbmix-8b-200000.pt |
| 220000 | step-220000 |
climbmix-8b-220000.pt |
| 240000 | step-240000 |
climbmix-8b-240000.pt |
| 250000 | step-250000 |
climbmix-8b-250000.pt |
| 260000 | step-260000 |
climbmix-8b-260000.pt |
| 280000 | step-280000 |
climbmix-8b-280000.pt |
| 300000 | step-300000 |
climbmix-8b-300000.pt |
| 320000 | step-320000 |
climbmix-8b-320000.pt |
| 340000 | step-340000 |
climbmix-8b-340000.pt |
| 360000 | step-360000 |
climbmix-8b-360000.pt |
Download one checkpoint
from huggingface_hub import hf_hub_download
import torch
path = hf_hub_download(
repo_id="daslab-testing/CloverLM-8b",
filename="climbmix-8b-20000.pt",
revision="step-20000", # pick the branch = training step you want
)
state = torch.load(path, map_location="cpu")
Or with the CLI:
hf download daslab-testing/CloverLM-8b climbmix-8b-20000.pt --revision step-20000 --local-dir .
Inference Providers NEW
This model isn't deployed by any Inference Provider. ๐ Ask for provider support