OLMo-150M-as_fm3_tg_omi1_omi2-k2: step58527-unsharded
Browse files- pretrain/OLMo-150M-as_fm3_tg_omi1_omi2-k2/step58527-unsharded/config.yaml +454 -0
- pretrain/OLMo-150M-as_fm3_tg_omi1_omi2-k2/step58527-unsharded/model.pt +3 -0
- pretrain/OLMo-150M-as_fm3_tg_omi1_omi2-k2/step58527-unsharded/optim.pt +3 -0
- pretrain/OLMo-150M-as_fm3_tg_omi1_omi2-k2/step58527-unsharded/train.pt +3 -0
pretrain/OLMo-150M-as_fm3_tg_omi1_omi2-k2/step58527-unsharded/config.yaml
ADDED
|
@@ -0,0 +1,454 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
run_name: OLMo-150M-as_fm3_tg_omi1_omi2-k2
|
| 2 |
+
seed: 6198
|
| 3 |
+
epoch: null
|
| 4 |
+
dry_run: false
|
| 5 |
+
model:
|
| 6 |
+
d_model: 768
|
| 7 |
+
n_heads: 12
|
| 8 |
+
n_kv_heads: null
|
| 9 |
+
clip_qkv: null
|
| 10 |
+
n_layers: 12
|
| 11 |
+
mlp_ratio: 8
|
| 12 |
+
mlp_hidden_size: null
|
| 13 |
+
activation_type: swiglu
|
| 14 |
+
block_type: sequential
|
| 15 |
+
block_group_size: 1
|
| 16 |
+
alibi: false
|
| 17 |
+
alibi_bias_max: 8.0
|
| 18 |
+
rope: true
|
| 19 |
+
rope_full_precision: true
|
| 20 |
+
rope_theta: 10000
|
| 21 |
+
flash_attention: true
|
| 22 |
+
attention_dropout: 0.0
|
| 23 |
+
multi_query_attention: null
|
| 24 |
+
attention_layer_norm: false
|
| 25 |
+
residual_dropout: 0.0
|
| 26 |
+
embedding_dropout: 0.0
|
| 27 |
+
embedding_layer_norm: false
|
| 28 |
+
layer_norm_type: default
|
| 29 |
+
layer_norm_with_affine: false
|
| 30 |
+
layer_norm_eps: 1.0e-05
|
| 31 |
+
attention_layer_norm_with_affine: false
|
| 32 |
+
max_sequence_length: 2048
|
| 33 |
+
include_bias: false
|
| 34 |
+
bias_for_layer_norm: false
|
| 35 |
+
scale_logits: false
|
| 36 |
+
vocab_size: 32000
|
| 37 |
+
embedding_size: 32000
|
| 38 |
+
weight_tying: false
|
| 39 |
+
eos_token_id: 0
|
| 40 |
+
pad_token_id: 1
|
| 41 |
+
init_device: cuda
|
| 42 |
+
init_fn: normal
|
| 43 |
+
init_std: 0.02
|
| 44 |
+
init_cutoff_factor: 3.0
|
| 45 |
+
precision: amp_bf16
|
| 46 |
+
scale_emb_init: false
|
| 47 |
+
emb_init_std: null
|
| 48 |
+
norm_after: false
|
| 49 |
+
optimizer:
|
| 50 |
+
name: adamw
|
| 51 |
+
learning_rate: 0.001
|
| 52 |
+
weight_decay: 0.1
|
| 53 |
+
betas:
|
| 54 |
+
- 0.9
|
| 55 |
+
- 0.95
|
| 56 |
+
eps: 1.0e-08
|
| 57 |
+
no_decay_norm_and_bias: null
|
| 58 |
+
selective_updates: false
|
| 59 |
+
decay_norm_and_bias: true
|
| 60 |
+
decay_embeddings: true
|
| 61 |
+
metrics_log_interval: 10
|
| 62 |
+
record_update_metrics: false
|
| 63 |
+
scheduler:
|
| 64 |
+
name: cosine_with_warmup
|
| 65 |
+
units: steps
|
| 66 |
+
t_warmup: 5000
|
| 67 |
+
t_max: null
|
| 68 |
+
alpha_f: 0.1
|
| 69 |
+
grad_clip_warmup_steps: null
|
| 70 |
+
grad_clip_warmup_factor: null
|
| 71 |
+
warmup_min_lr: 0.0
|
| 72 |
+
data:
|
| 73 |
+
paths:
|
| 74 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/algebraic-stack-tokenized/00000_00000_shuffled.ds
|
| 75 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/algebraic-stack-tokenized/00001_00000_shuffled.ds
|
| 76 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/algebraic-stack-tokenized/00002_00000_shuffled.ds
|
| 77 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/algebraic-stack-tokenized/00003_00000_shuffled.ds
|
| 78 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/algebraic-stack-tokenized/00004_00000_shuffled.ds
|
| 79 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/algebraic-stack-tokenized/00005_00000_shuffled.ds
|
| 80 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/algebraic-stack-tokenized/00006_00000_shuffled.ds
|
| 81 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/algebraic-stack-tokenized/00007_00000_shuffled.ds
|
| 82 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/algebraic-stack-tokenized/00008_00000_shuffled.ds
|
| 83 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/algebraic-stack-tokenized/00009_00000_shuffled.ds
|
| 84 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/algebraic-stack-tokenized/00010_00000_shuffled.ds
|
| 85 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/algebraic-stack-tokenized/00011_00000_shuffled.ds
|
| 86 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/algebraic-stack-tokenized/00012_00000_shuffled.ds
|
| 87 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/algebraic-stack-tokenized/00013_00000_shuffled.ds
|
| 88 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/algebraic-stack-tokenized/00014_00000_shuffled.ds
|
| 89 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/algebraic-stack-tokenized/00016_00000_shuffled.ds
|
| 90 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/algebraic-stack-tokenized/00019_00000_shuffled.ds
|
| 91 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/algebraic-stack-tokenized/00020_00000_shuffled.ds
|
| 92 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/algebraic-stack-tokenized/00021_00000_shuffled.ds
|
| 93 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/algebraic-stack-tokenized/00022_00000_shuffled.ds
|
| 94 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/algebraic-stack-tokenized/00023_00000_shuffled.ds
|
| 95 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/algebraic-stack-tokenized/00024_00000_shuffled.ds
|
| 96 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/algebraic-stack-tokenized/00025_00000_shuffled.ds
|
| 97 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/algebraic-stack-tokenized/00026_00000_shuffled.ds
|
| 98 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/algebraic-stack-tokenized/00027_00000_shuffled.ds
|
| 99 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/algebraic-stack-tokenized/00028_00000_shuffled.ds
|
| 100 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/algebraic-stack-tokenized/00029_00000_shuffled.ds
|
| 101 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/algebraic-stack-tokenized/00030_00000_shuffled.ds
|
| 102 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/algebraic-stack-tokenized/00031_00000_shuffled.ds
|
| 103 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/algebraic-stack-tokenized/00032_00000_shuffled.ds
|
| 104 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/algebraic-stack-tokenized/00033_00000_shuffled.ds
|
| 105 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/algebraic-stack-tokenized/00034_00000_shuffled.ds
|
| 106 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/algebraic-stack-tokenized/00035_00000_shuffled.ds
|
| 107 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/algebraic-stack-tokenized/00036_00000_shuffled.ds
|
| 108 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/algebraic-stack-tokenized/00037_00000_shuffled.ds
|
| 109 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/algebraic-stack-tokenized/00038_00000_shuffled.ds
|
| 110 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/algebraic-stack-tokenized/00039_00000_shuffled.ds
|
| 111 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/algebraic-stack-tokenized/00040_00000_shuffled.ds
|
| 112 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/algebraic-stack-tokenized/00041_00000_shuffled.ds
|
| 113 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/algebraic-stack-tokenized/00042_00000_shuffled.ds
|
| 114 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/algebraic-stack-tokenized/00043_00000_shuffled.ds
|
| 115 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/algebraic-stack-tokenized/00044_00000_shuffled.ds
|
| 116 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/algebraic-stack-tokenized/00045_00000_shuffled.ds
|
| 117 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/algebraic-stack-tokenized/00046_00000_shuffled.ds
|
| 118 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/algebraic-stack-tokenized/00047_00000_shuffled.ds
|
| 119 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/algebraic-stack-tokenized/00048_00000_shuffled.ds
|
| 120 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/algebraic-stack-tokenized/00049_00000_shuffled.ds
|
| 121 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/algebraic-stack-tokenized/00050_00000_shuffled.ds
|
| 122 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/algebraic-stack-tokenized/00051_00000_shuffled.ds
|
| 123 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/algebraic-stack-tokenized/00052_00000_shuffled.ds
|
| 124 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/algebraic-stack-tokenized/00053_00000_shuffled.ds
|
| 125 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/algebraic-stack-tokenized/00054_00000_shuffled.ds
|
| 126 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/algebraic-stack-tokenized/00055_00000_shuffled.ds
|
| 127 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/algebraic-stack-tokenized/00056_00000_shuffled.ds
|
| 128 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/algebraic-stack-tokenized/00057_00000_shuffled.ds
|
| 129 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/algebraic-stack-tokenized/00058_00000_shuffled.ds
|
| 130 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/algebraic-stack-tokenized/00059_00000_shuffled.ds
|
| 131 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/algebraic-stack-tokenized/00060_00000_shuffled.ds
|
| 132 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/algebraic-stack-tokenized/00061_00000_shuffled.ds
|
| 133 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/algebraic-stack-tokenized/00062_00000_shuffled.ds
|
| 134 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/algebraic-stack-tokenized/00063_00000_shuffled.ds
|
| 135 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/finemath3-tokenized/00000_00000_shuffled.ds
|
| 136 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/finemath3-tokenized/00001_00000_shuffled.ds
|
| 137 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/finemath3-tokenized/00002_00000_shuffled.ds
|
| 138 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/finemath3-tokenized/00003_00000_shuffled.ds
|
| 139 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/finemath3-tokenized/00004_00000_shuffled.ds
|
| 140 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/finemath3-tokenized/00005_00000_shuffled.ds
|
| 141 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/finemath3-tokenized/00006_00000_shuffled.ds
|
| 142 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/finemath3-tokenized/00007_00000_shuffled.ds
|
| 143 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/finemath3-tokenized/00008_00000_shuffled.ds
|
| 144 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/finemath3-tokenized/00009_00000_shuffled.ds
|
| 145 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/finemath3-tokenized/00010_00000_shuffled.ds
|
| 146 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/finemath3-tokenized/00011_00000_shuffled.ds
|
| 147 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/finemath3-tokenized/00012_00000_shuffled.ds
|
| 148 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/finemath3-tokenized/00013_00000_shuffled.ds
|
| 149 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/finemath3-tokenized/00014_00000_shuffled.ds
|
| 150 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/finemath3-tokenized/00015_00000_shuffled.ds
|
| 151 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/finemath3-tokenized/00016_00000_shuffled.ds
|
| 152 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/finemath3-tokenized/00017_00000_shuffled.ds
|
| 153 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/finemath3-tokenized/00018_00000_shuffled.ds
|
| 154 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/finemath3-tokenized/00019_00000_shuffled.ds
|
| 155 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/finemath3-tokenized/00020_00000_shuffled.ds
|
| 156 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/finemath3-tokenized/00021_00000_shuffled.ds
|
| 157 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/finemath3-tokenized/00022_00000_shuffled.ds
|
| 158 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/finemath3-tokenized/00023_00000_shuffled.ds
|
| 159 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/finemath3-tokenized/00024_00000_shuffled.ds
|
| 160 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/finemath3-tokenized/00025_00000_shuffled.ds
|
| 161 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/finemath3-tokenized/00026_00000_shuffled.ds
|
| 162 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/finemath3-tokenized/00027_00000_shuffled.ds
|
| 163 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/finemath3-tokenized/00028_00000_shuffled.ds
|
| 164 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/finemath3-tokenized/00029_00000_shuffled.ds
|
| 165 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/finemath3-tokenized/00030_00000_shuffled.ds
|
| 166 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/finemath3-tokenized/00031_00000_shuffled.ds
|
| 167 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/finemath3-tokenized/00032_00000_shuffled.ds
|
| 168 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/finemath3-tokenized/00033_00000_shuffled.ds
|
| 169 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/finemath3-tokenized/00034_00000_shuffled.ds
|
| 170 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/finemath3-tokenized/00035_00000_shuffled.ds
|
| 171 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/finemath3-tokenized/00036_00000_shuffled.ds
|
| 172 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/finemath3-tokenized/00037_00000_shuffled.ds
|
| 173 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/finemath3-tokenized/00038_00000_shuffled.ds
|
| 174 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/finemath3-tokenized/00039_00000_shuffled.ds
|
| 175 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/finemath3-tokenized/00040_00000_shuffled.ds
|
| 176 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/finemath3-tokenized/00041_00000_shuffled.ds
|
| 177 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/finemath3-tokenized/00042_00000_shuffled.ds
|
| 178 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/finemath3-tokenized/00043_00000_shuffled.ds
|
| 179 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/finemath3-tokenized/00044_00000_shuffled.ds
|
| 180 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/finemath3-tokenized/00045_00000_shuffled.ds
|
| 181 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/finemath3-tokenized/00046_00000_shuffled.ds
|
| 182 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/finemath3-tokenized/00047_00000_shuffled.ds
|
| 183 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/finemath3-tokenized/00048_00000_shuffled.ds
|
| 184 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/finemath3-tokenized/00049_00000_shuffled.ds
|
| 185 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/finemath3-tokenized/00050_00000_shuffled.ds
|
| 186 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/finemath3-tokenized/00051_00000_shuffled.ds
|
| 187 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/finemath3-tokenized/00052_00000_shuffled.ds
|
| 188 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/finemath3-tokenized/00053_00000_shuffled.ds
|
| 189 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/finemath3-tokenized/00054_00000_shuffled.ds
|
| 190 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/finemath3-tokenized/00055_00000_shuffled.ds
|
| 191 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/finemath3-tokenized/00056_00000_shuffled.ds
|
| 192 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/finemath3-tokenized/00057_00000_shuffled.ds
|
| 193 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/finemath3-tokenized/00058_00000_shuffled.ds
|
| 194 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/finemath3-tokenized/00059_00000_shuffled.ds
|
| 195 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/finemath3-tokenized/00060_00000_shuffled.ds
|
| 196 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/finemath3-tokenized/00061_00000_shuffled.ds
|
| 197 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/finemath3-tokenized/00062_00000_shuffled.ds
|
| 198 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/finemath3-tokenized/00063_00000_shuffled.ds
|
| 199 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/openmathinstruct1-tokenized/00000_00000_shuffled.ds
|
| 200 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/openmathinstruct2-tokenized/00000_00000_shuffled.ds
|
| 201 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/openmathinstruct2-tokenized/00001_00000_shuffled.ds
|
| 202 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/openmathinstruct2-tokenized/00002_00000_shuffled.ds
|
| 203 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/openmathinstruct2-tokenized/00003_00000_shuffled.ds
|
| 204 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/openmathinstruct2-tokenized/00004_00000_shuffled.ds
|
| 205 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/openmathinstruct2-tokenized/00005_00000_shuffled.ds
|
| 206 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/openmathinstruct2-tokenized/00006_00000_shuffled.ds
|
| 207 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/openmathinstruct2-tokenized/00007_00000_shuffled.ds
|
| 208 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/openmathinstruct2-tokenized/00008_00000_shuffled.ds
|
| 209 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/openmathinstruct2-tokenized/00009_00000_shuffled.ds
|
| 210 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/openmathinstruct2-tokenized/00010_00000_shuffled.ds
|
| 211 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/openmathinstruct2-tokenized/00011_00000_shuffled.ds
|
| 212 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/openmathinstruct2-tokenized/00012_00000_shuffled.ds
|
| 213 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/openmathinstruct2-tokenized/00013_00000_shuffled.ds
|
| 214 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/openmathinstruct2-tokenized/00014_00000_shuffled.ds
|
| 215 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/openmathinstruct2-tokenized/00015_00000_shuffled.ds
|
| 216 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/openmathinstruct2-tokenized/00016_00000_shuffled.ds
|
| 217 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/openmathinstruct2-tokenized/00017_00000_shuffled.ds
|
| 218 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/openmathinstruct2-tokenized/00018_00000_shuffled.ds
|
| 219 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/openmathinstruct2-tokenized/00019_00000_shuffled.ds
|
| 220 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/openmathinstruct2-tokenized/00020_00000_shuffled.ds
|
| 221 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/openmathinstruct2-tokenized/00021_00000_shuffled.ds
|
| 222 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/openmathinstruct2-tokenized/00022_00000_shuffled.ds
|
| 223 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/openmathinstruct2-tokenized/00023_00000_shuffled.ds
|
| 224 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/openmathinstruct2-tokenized/00024_00000_shuffled.ds
|
| 225 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/openmathinstruct2-tokenized/00025_00000_shuffled.ds
|
| 226 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/openmathinstruct2-tokenized/00026_00000_shuffled.ds
|
| 227 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/openmathinstruct2-tokenized/00027_00000_shuffled.ds
|
| 228 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/openmathinstruct2-tokenized/00028_00000_shuffled.ds
|
| 229 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/openmathinstruct2-tokenized/00029_00000_shuffled.ds
|
| 230 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/openmathinstruct2-tokenized/00030_00000_shuffled.ds
|
| 231 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/openmathinstruct2-tokenized/00031_00000_shuffled.ds
|
| 232 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/tinygsm-tokenized/00000_00000_shuffled.ds
|
| 233 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/tinygsm-tokenized/00001_00000_shuffled.ds
|
| 234 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/tinygsm-tokenized/00002_00000_shuffled.ds
|
| 235 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/tinygsm-tokenized/00003_00000_shuffled.ds
|
| 236 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/tinygsm-tokenized/00004_00000_shuffled.ds
|
| 237 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/tinygsm-tokenized/00005_00000_shuffled.ds
|
| 238 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/tinygsm-tokenized/00006_00000_shuffled.ds
|
| 239 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/tinygsm-tokenized/00007_00000_shuffled.ds
|
| 240 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/tinygsm-tokenized/00008_00000_shuffled.ds
|
| 241 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/tinygsm-tokenized/00009_00000_shuffled.ds
|
| 242 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/tinygsm-tokenized/00010_00000_shuffled.ds
|
| 243 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/tinygsm-tokenized/00011_00000_shuffled.ds
|
| 244 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/tinygsm-tokenized/00012_00000_shuffled.ds
|
| 245 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/tinygsm-tokenized/00013_00000_shuffled.ds
|
| 246 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/tinygsm-tokenized/00014_00000_shuffled.ds
|
| 247 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/tinygsm-tokenized/00015_00000_shuffled.ds
|
| 248 |
+
- /data/home/guanning/olmo_data/mathonly/mix/as_fm3_tg_omi1_omi2/tinygsm-tokenized/00016_00000_shuffled.ds
|
| 249 |
+
memmap_dtype: uint16
|
| 250 |
+
datasets: null
|
| 251 |
+
label_mask_paths: null
|
| 252 |
+
pad_direction: right
|
| 253 |
+
generate_attention_mask: false
|
| 254 |
+
generate_doc_lengths: false
|
| 255 |
+
num_workers: 32
|
| 256 |
+
drop_last: true
|
| 257 |
+
pin_memory: true
|
| 258 |
+
prefetch_factor: 8
|
| 259 |
+
persistent_workers: true
|
| 260 |
+
timeout: 0
|
| 261 |
+
seed: null
|
| 262 |
+
instance_filter: null
|
| 263 |
+
custom_dataset: null
|
| 264 |
+
restore_dataloader: true
|
| 265 |
+
fast_forward_batches: null
|
| 266 |
+
evaluators:
|
| 267 |
+
- label: gsm8k-test
|
| 268 |
+
type: lm
|
| 269 |
+
data:
|
| 270 |
+
paths:
|
| 271 |
+
- /data/home/guanning/olmo_data/mathonly/eval/gsm8k-test-tokenized/00000_00000_shuffled.ds
|
| 272 |
+
memmap_dtype: uint16
|
| 273 |
+
datasets: null
|
| 274 |
+
label_mask_paths: null
|
| 275 |
+
pad_direction: right
|
| 276 |
+
generate_attention_mask: false
|
| 277 |
+
generate_doc_lengths: false
|
| 278 |
+
num_workers: 0
|
| 279 |
+
drop_last: true
|
| 280 |
+
pin_memory: false
|
| 281 |
+
prefetch_factor: null
|
| 282 |
+
persistent_workers: false
|
| 283 |
+
timeout: 0
|
| 284 |
+
seed: null
|
| 285 |
+
instance_filter: null
|
| 286 |
+
custom_dataset: null
|
| 287 |
+
device_eval_batch_size: null
|
| 288 |
+
subset_num_batches: null
|
| 289 |
+
- label: math500
|
| 290 |
+
type: lm
|
| 291 |
+
data:
|
| 292 |
+
paths:
|
| 293 |
+
- /data/home/guanning/olmo_data/mathonly/eval/math500-tokenized/00000_00000_shuffled.ds
|
| 294 |
+
memmap_dtype: uint16
|
| 295 |
+
datasets: null
|
| 296 |
+
label_mask_paths: null
|
| 297 |
+
pad_direction: right
|
| 298 |
+
generate_attention_mask: false
|
| 299 |
+
generate_doc_lengths: false
|
| 300 |
+
num_workers: 0
|
| 301 |
+
drop_last: true
|
| 302 |
+
pin_memory: false
|
| 303 |
+
prefetch_factor: null
|
| 304 |
+
persistent_workers: false
|
| 305 |
+
timeout: 0
|
| 306 |
+
seed: null
|
| 307 |
+
instance_filter: null
|
| 308 |
+
custom_dataset: null
|
| 309 |
+
device_eval_batch_size: null
|
| 310 |
+
subset_num_batches: null
|
| 311 |
+
- label: openmathinstruct1-validation
|
| 312 |
+
type: lm
|
| 313 |
+
data:
|
| 314 |
+
paths:
|
| 315 |
+
- /data/home/guanning/olmo_data/mathonly/eval/openmathinstruct1-validation-tokenized/00000_00000_shuffled.ds
|
| 316 |
+
memmap_dtype: uint16
|
| 317 |
+
datasets: null
|
| 318 |
+
label_mask_paths: null
|
| 319 |
+
pad_direction: right
|
| 320 |
+
generate_attention_mask: false
|
| 321 |
+
generate_doc_lengths: false
|
| 322 |
+
num_workers: 0
|
| 323 |
+
drop_last: true
|
| 324 |
+
pin_memory: false
|
| 325 |
+
prefetch_factor: null
|
| 326 |
+
persistent_workers: false
|
| 327 |
+
timeout: 0
|
| 328 |
+
seed: null
|
| 329 |
+
instance_filter: null
|
| 330 |
+
custom_dataset: null
|
| 331 |
+
device_eval_batch_size: null
|
| 332 |
+
subset_num_batches: 50
|
| 333 |
+
- label: gsm8k_gold_bpb_5shot
|
| 334 |
+
type: downstream
|
| 335 |
+
data:
|
| 336 |
+
paths: null
|
| 337 |
+
memmap_dtype: uint16
|
| 338 |
+
datasets: null
|
| 339 |
+
label_mask_paths: null
|
| 340 |
+
pad_direction: right
|
| 341 |
+
generate_attention_mask: false
|
| 342 |
+
generate_doc_lengths: false
|
| 343 |
+
num_workers: 0
|
| 344 |
+
drop_last: false
|
| 345 |
+
pin_memory: false
|
| 346 |
+
prefetch_factor: null
|
| 347 |
+
persistent_workers: false
|
| 348 |
+
timeout: 0
|
| 349 |
+
seed: null
|
| 350 |
+
instance_filter: null
|
| 351 |
+
custom_dataset: null
|
| 352 |
+
device_eval_batch_size: null
|
| 353 |
+
subset_num_batches: null
|
| 354 |
+
- label: basic_arithmetic
|
| 355 |
+
type: downstream
|
| 356 |
+
data:
|
| 357 |
+
paths: null
|
| 358 |
+
memmap_dtype: uint16
|
| 359 |
+
datasets: null
|
| 360 |
+
label_mask_paths: null
|
| 361 |
+
pad_direction: right
|
| 362 |
+
generate_attention_mask: false
|
| 363 |
+
generate_doc_lengths: false
|
| 364 |
+
num_workers: 0
|
| 365 |
+
drop_last: false
|
| 366 |
+
pin_memory: false
|
| 367 |
+
prefetch_factor: null
|
| 368 |
+
persistent_workers: false
|
| 369 |
+
timeout: 0
|
| 370 |
+
seed: null
|
| 371 |
+
instance_filter: null
|
| 372 |
+
custom_dataset: null
|
| 373 |
+
device_eval_batch_size: null
|
| 374 |
+
subset_num_batches: null
|
| 375 |
+
eval_interval: 1000
|
| 376 |
+
tokenizer:
|
| 377 |
+
identifier: rosieyzh/OLMo-150M-4xtg
|
| 378 |
+
truncate_direction: right
|
| 379 |
+
save_folder: /data/home/guanning/olmo_runs/mathonly/OLMo-150M-as_fm3_tg_omi1_omi2-k2
|
| 380 |
+
remote_save_folder: null
|
| 381 |
+
canceled_check_interval: 50
|
| 382 |
+
save_interval: 2000
|
| 383 |
+
save_interval_unsharded: 5000
|
| 384 |
+
save_interval_ephemeral: null
|
| 385 |
+
save_num_checkpoints_to_keep: 1
|
| 386 |
+
save_num_unsharded_checkpoints_to_keep: 2
|
| 387 |
+
save_overwrite: true
|
| 388 |
+
force_save_unsharded: false
|
| 389 |
+
no_pre_train_checkpoint: false
|
| 390 |
+
load_path: null
|
| 391 |
+
load_path_sharded_checkpointer: null
|
| 392 |
+
try_load_latest_save: true
|
| 393 |
+
reset_optimizer_state: false
|
| 394 |
+
reset_trainer_state: false
|
| 395 |
+
sharded_checkpointer: torch_legacy
|
| 396 |
+
new_style_checkpoints: null
|
| 397 |
+
max_duration: 1ep
|
| 398 |
+
global_train_batch_size: 512
|
| 399 |
+
device_train_batch_size: 64
|
| 400 |
+
device_train_microbatch_size: 16
|
| 401 |
+
device_eval_batch_size: 16
|
| 402 |
+
eval_subset_num_batches: -1
|
| 403 |
+
eval_on_load: false
|
| 404 |
+
device_train_grad_accum: 4
|
| 405 |
+
max_grad_norm: 1.0
|
| 406 |
+
max_grad_norm_ratio: null
|
| 407 |
+
precision: amp_bf16
|
| 408 |
+
wandb:
|
| 409 |
+
project: olmo-mathonly-1018
|
| 410 |
+
entity: train-llm
|
| 411 |
+
group: OLMo-150M
|
| 412 |
+
name: OLMo-150M-as_fm3_tg_omi1_omi2-k2
|
| 413 |
+
tags:
|
| 414 |
+
- mathonly
|
| 415 |
+
- 150M
|
| 416 |
+
- k2
|
| 417 |
+
log_artifacts: false
|
| 418 |
+
rank_zero_only: true
|
| 419 |
+
log_interval: 1
|
| 420 |
+
speed_monitor:
|
| 421 |
+
window_size: 20
|
| 422 |
+
gpu_flops_available: null
|
| 423 |
+
console_log_interval: 1
|
| 424 |
+
gen1_gc_interval: 1
|
| 425 |
+
compile: null
|
| 426 |
+
distributed_strategy: ddp
|
| 427 |
+
fsdp:
|
| 428 |
+
use_orig_params: true
|
| 429 |
+
sharding_strategy: FULL_SHARD
|
| 430 |
+
wrapping_strategy: null
|
| 431 |
+
precision: pure
|
| 432 |
+
hybrid_sharding_num_model_replicas: null
|
| 433 |
+
ddp:
|
| 434 |
+
grad_sync_mode: batch
|
| 435 |
+
find_unused_params: false
|
| 436 |
+
single:
|
| 437 |
+
device: auto
|
| 438 |
+
softmax_auxiliary_loss: false
|
| 439 |
+
auxiliary_loss_multiplier: 0.0001
|
| 440 |
+
passk_k: 2
|
| 441 |
+
passk_mix: 0
|
| 442 |
+
passk_log_top: 65536
|
| 443 |
+
time_limit: null
|
| 444 |
+
extra_steps_after_cancel: 10
|
| 445 |
+
early_stopping_factor: null
|
| 446 |
+
save_data_indices: true
|
| 447 |
+
python_profiling: false
|
| 448 |
+
torch_profiling: false
|
| 449 |
+
stop_at: 58537
|
| 450 |
+
stop_after: null
|
| 451 |
+
activation_checkpointing: null
|
| 452 |
+
fused_loss: null
|
| 453 |
+
hf_datasets_cache_dir: null
|
| 454 |
+
module_outputs_save_steps: null
|
pretrain/OLMo-150M-as_fm3_tg_omi1_omi2-k2/step58527-unsharded/model.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b10769a11b5fb5f9674f002876464e7c9b6836d8aa464a21d00a8e726dd78f0d
|
| 3 |
+
size 649612628
|
pretrain/OLMo-150M-as_fm3_tg_omi1_omi2-k2/step58527-unsharded/optim.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:38ed07681a60da3d732f71df9f543c82944b1dc7aecca3a6d35dcf4146f29e46
|
| 3 |
+
size 1299223890
|
pretrain/OLMo-150M-as_fm3_tg_omi1_omi2-k2/step58527-unsharded/train.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e3ac569bc6a1d030a457f562a4a3f27931eb7b724cef1db00a60663d4bd48e52
|
| 3 |
+
size 14732
|