Downloads · 30 days
30
7% of all-time downloads
deepseek70b-7816-merged-bf16
koreallmdev/deepseek70b-7816-merged-bf16
deepseek70b-7816-merged-bf16 is a text generation model from koreallmdev. Use it when you need the model to write or continue text. It is set up for transformers. The card lists the license as mit.
Standalone BF16 model produced by merging the Final 7,816-row LoRA adapter into deepseek-ai/DeepSeek-R1-Distill-Llama-70B.
All-time downloads
405
Public
Parameters
70.6B
141 GB on disk
Likes
1
Public
Repo makeup
Click a slice to open those files.
.safetensors141 GB · 100%
At a glance
- Task
- Text Generation
- Library
- transformers
- License
- mit
- Model type
- llama
- Access
- Public
- Created
- Aug 4, 2026
- Updated
- Aug 4, 2026
- SHA
- 2d6c8db2
Try a prompt
Base models
Model card
From the Hugging Face model README
DeepSeek 70B Final 7,816 — Standalone Merged BF16
Standalone BF16 model produced by merging the Final 7,816-row LoRA adapter into
deepseek-ai/DeepSeek-R1-Distill-Llama-70B.
- Base model license: MIT
- Unique training rows: 7,816
- Held-out Eval rows: 200
- Adapter required after download: No
- Merge method: memory-safe safetensors shard-by-shard LoRA delta merge
- Release status: Conditional Stable
- Zero-defect claim: No
The repository includes MERGE_REPORT.json, the original configuration and
tokenizer, a shard index, checksums, and standalone Smoke3 evidence.
- Task
- Text Generation
- Library
- transformers
- Type
- llama
- License
- mit
- Created
- Aug 4, 2026
- Updated
- Aug 4, 2026