Jonathan von Rad

jvonrad's activity

• All
• Models
• Datasets
• Spaces
• Buckets
• Papers
• Collections
• Community
• Posts
• Upvotes
• Likes
• Articles

New activity in hugging-quants/Meta-Llama-3.1-8B-Instruct-AWQ-INT4 4 months ago

Tip: For Hardware Acceleration this Model will not leverage vllm marlin kernels!

#19 opened 4 months ago by

jvonrad

New activity in Qwen/Qwen3-32B 5 months ago

Where is the Base Model?

👍➕ 10

3

#34 opened 11 months ago by

jvonrad

New activity in Harvard-DCML/boomerang-qwen3-4.9B 5 months ago

Substantially lower accuracy on reasoning benchmarks such as GSM8K (1.5%) and MATH-500 (4.2%)

1

#1 opened 5 months ago by

jvonrad

New activity in monology/pile-uncopyrighted 10 months ago

Could you please implement train:1% feature? This way we don't have to download the entire dataset.

1

#12 opened 10 months ago by

jvonrad

New activity in Qwen/Qwen3-32B 10 months ago

Low Score on GSM8K on lm-eval-harness? (just 74.91)

2

#36 opened 10 months ago by

jvonrad

New activity in nvidia/NV-Embed-v2 10 months ago

TypeError: cannot unpack non-iterable NoneType object

👀👍 8

5

#37 opened over 1 year ago by

Pietroferr

New activity in google/gemma-3-27b-pt 11 months ago

Model is a Memory Hog - 2xH100 80GB OOM??

1

#5 opened 11 months ago by

jvonrad

New activity in google/gemma-3-1b-pt 11 months ago

When evaluating Wiki2, I just get Loss: Nan, while with gemma-3-1b-it it works..

2

#8 opened 11 months ago by

jvonrad