Wing Lian 22810c97b7 use warmup_ratio as a better default than warmup steps since it's data dependent (#2897) [skip ci] ...

* use warmup_ratio as a better default than warmup steps since it's data dependent

* replace remainder of warmup_steps

2025-07-30 06:44:06 -04:00

..

alst

TiledMLP support for FSDP2 (#2950)

2025-07-25 07:15:03 -04:00

archived

use warmup_ratio as a better default than warmup steps since it's data dependent (#2897) [skip ci]

2025-07-30 06:44:06 -04:00

cloud

make pad_to_sequence_len default to the same value as sample_packing (#2941) [skip ci]

2025-07-21 11:40:56 -04:00

cohere

make pad_to_sequence_len default to the same value as sample_packing (#2941) [skip ci]

2025-07-21 11:40:56 -04:00

colab-notebooks

Feat: Add voxtral, magistral small 1.1, and misc gemma3n fixes (#2979)

2025-07-30 15:57:05 +07:00

deepcogito

use warmup_ratio as a better default than warmup steps since it's data dependent (#2897) [skip ci]

2025-07-30 06:44:06 -04:00

deepseek-v2

use warmup_ratio as a better default than warmup steps since it's data dependent (#2897) [skip ci]

2025-07-30 06:44:06 -04:00

devstral

make pad_to_sequence_len default to the same value as sample_packing (#2941) [skip ci]

2025-07-21 11:40:56 -04:00

falcon-h1

make pad_to_sequence_len default to the same value as sample_packing (#2941) [skip ci]

2025-07-21 11:40:56 -04:00

gemma2

make pad_to_sequence_len default to the same value as sample_packing (#2941) [skip ci]

2025-07-21 11:40:56 -04:00

gemma3

make pad_to_sequence_len default to the same value as sample_packing (#2941) [skip ci]

2025-07-21 11:40:56 -04:00

gemma3n

Feat: Add voxtral, magistral small 1.1, and misc gemma3n fixes (#2979)

2025-07-30 15:57:05 +07:00

glm4

use warmup_ratio as a better default than warmup steps since it's data dependent (#2897) [skip ci]

2025-07-30 06:44:06 -04:00

jamba

use warmup_ratio as a better default than warmup steps since it's data dependent (#2897) [skip ci]

2025-07-30 06:44:06 -04:00

lfm2

make pad_to_sequence_len default to the same value as sample_packing (#2941) [skip ci]

2025-07-21 11:40:56 -04:00

llama-2

use warmup_ratio as a better default than warmup steps since it's data dependent (#2897) [skip ci]

2025-07-30 06:44:06 -04:00

llama-3

use warmup_ratio as a better default than warmup steps since it's data dependent (#2897) [skip ci]

2025-07-30 06:44:06 -04:00

llama-3-vision

fix: revert changing default optimizer to muon (#2965) [skip ci]

2025-07-22 10:00:30 -04:00

llama-4

use warmup_ratio as a better default than warmup steps since it's data dependent (#2897) [skip ci]

2025-07-30 06:44:06 -04:00

llava

fix: revert changing default optimizer to muon (#2965) [skip ci]

2025-07-22 10:00:30 -04:00

magistral

Feat: Add voxtral, magistral small 1.1, and misc gemma3n fixes (#2979)

2025-07-30 15:57:05 +07:00

mamba

use warmup_ratio as a better default than warmup steps since it's data dependent (#2897) [skip ci]

2025-07-30 06:44:06 -04:00

mistral

use warmup_ratio as a better default than warmup steps since it's data dependent (#2897) [skip ci]

2025-07-30 06:44:06 -04:00

orpheus

use warmup_ratio as a better default than warmup steps since it's data dependent (#2897) [skip ci]

2025-07-30 06:44:06 -04:00

phi

use warmup_ratio as a better default than warmup steps since it's data dependent (#2897) [skip ci]

2025-07-30 06:44:06 -04:00

pixtral

fix: revert changing default optimizer to muon (#2965) [skip ci]

2025-07-22 10:00:30 -04:00

qwen2

use warmup_ratio as a better default than warmup steps since it's data dependent (#2897) [skip ci]

2025-07-30 06:44:06 -04:00

qwen2_5-vl

checkpoint model on first step callback (#2906)

2025-07-15 15:00:48 -04:00

qwen2-vl

checkpoint model on first step callback (#2906)

2025-07-15 15:00:48 -04:00

qwen3

use warmup_ratio as a better default than warmup steps since it's data dependent (#2897) [skip ci]

2025-07-30 06:44:06 -04:00

voxtral

Feat: Add voxtral, magistral small 1.1, and misc gemma3n fixes (#2979)

2025-07-30 15:57:05 +07:00