Search results

qwen3-0.6b fine_tuning 64k OOM

Status: Open.

#67 In jquesnelle/yarn;

· dgzxx-2000 opened on Jun 9, 2025

ModuleNotFoundError: No module named 'scaled_rope.modeling_qwen3_yarn'

Status: Open.

#66 In jquesnelle/yarn;

· dgzxx-2000 opened on Jun 9, 2025

Have you tried Rope_base = 500k in YaRN?

Status: Open.

#64 In jquesnelle/yarn;

· hannlp opened on Sep 4, 2024

For models with other architectures, such as Qwen family, how to find the best <code>\alpha</code>, <code>\beta</code> and <code>\sqrt{1/t}</code> parameters?

Status: Open.

#63 In jquesnelle/yarn;

· ki-ljl opened on Jul 3, 2024

the detail of code implements.

Status: Open.

#62 In jquesnelle/yarn;

· cooper12121 opened on Jun 14, 2024

Question related to _yarn_linear_ramp_mask

Status: Open.

#60 In jquesnelle/yarn;

· chizhang118 opened on Apr 9, 2024

OOM error of distributed training on 80GB GPUs with Mistral-7b

Status: Open.

#59 In jquesnelle/yarn;

· TracyPlus opened on Apr 8, 2024

cannot connect to hugging face

Status: Open.

#58 In jquesnelle/yarn;

· RENNY-Jenius opened on Mar 29, 2024

Why the updated cache is initialized with seqlen=256?

Status: Open.

#57 In jquesnelle/yarn;

· pUmpKin-Co opened on Mar 27, 2024

Trying to set a tensor of shape torch.Size([257, 1024]) in "weight" (which has shape torch.Size([1226, 1024])), this look incorrect

Status: Open.

#56 In jquesnelle/yarn;

· Litong-sTs opened on Mar 18, 2024

Can we run the replication of the results，8 * 80 A100

Status: Open.

#55 In jquesnelle/yarn;

· zhanglv0209 opened on Mar 15, 2024

How should I proceed with conducting an evaluation for lm-evaluation-harness?

Status: Open.

#54 In jquesnelle/yarn;

· yuri-son opened on Mar 15, 2024