8B-tulu-sft-int2 / quant_log.csv
Matt300209's picture
Upload folder using huggingface_hub
de430e1 verified
raw
history blame
5.02 kB
layer,module,loss,samples,damp,time
0,self_attn.k_proj,8.79510689,0.01000,0.659
0,self_attn.v_proj,0.21052508,0.01000,0.466
0,self_attn.q_proj,18.25806046,0.01000,0.473
0,self_attn.o_proj,0.02193195,0.01000,0.470
0,mlp.up_proj,15.79567146,0.01000,0.476
0,mlp.gate_proj,20.15350723,0.01000,0.474
0,mlp.down_proj,0.15333621,0.01000,2.037
1,self_attn.k_proj,18.12412453,0.01000,0.472
1,self_attn.v_proj,0.98016936,0.01000,0.471
1,self_attn.q_proj,33.08531570,0.01000,0.476
1,self_attn.o_proj,0.11857565,0.01000,0.472
1,mlp.up_proj,24.87150764,0.01000,0.482
1,mlp.gate_proj,34.42536926,0.01000,0.479
1,mlp.down_proj,20.08703995,0.01000,1.996
2,self_attn.k_proj,28.00672913,0.01000,0.469
2,self_attn.v_proj,1.93891215,0.01000,0.468
2,self_attn.q_proj,55.34558868,0.01000,0.478
2,self_attn.o_proj,0.09205560,0.01000,0.470
2,mlp.up_proj,30.34683800,0.01000,0.479
2,mlp.gate_proj,49.06833267,0.01000,0.501
2,mlp.down_proj,0.42902470,0.01000,2.014
3,self_attn.k_proj,21.30574417,0.01000,0.473
3,self_attn.v_proj,2.84672379,0.01000,0.469
3,self_attn.q_proj,46.99731827,0.01000,0.474
3,self_attn.o_proj,0.16826847,0.01000,0.472
3,mlp.up_proj,38.18422699,0.01000,0.479
3,mlp.gate_proj,76.93240356,0.01000,0.477
3,mlp.down_proj,0.58045858,0.01000,1.992
4,self_attn.k_proj,24.07534409,0.01000,0.467
4,self_attn.v_proj,2.88128042,0.01000,0.468
4,self_attn.q_proj,50.19567108,0.01000,0.471
4,self_attn.o_proj,0.22788243,0.01000,0.471
4,mlp.up_proj,40.49008560,0.01000,0.480
4,mlp.gate_proj,88.19657898,0.01000,0.475
4,mlp.down_proj,0.64514506,0.01000,1.998
5,self_attn.k_proj,35.62877274,0.01000,0.473
5,self_attn.v_proj,2.56014800,0.01000,0.469
5,self_attn.q_proj,64.17327118,0.01000,0.478
5,self_attn.o_proj,0.23322049,0.01000,0.507
5,mlp.up_proj,42.68757629,0.01000,0.481
5,mlp.gate_proj,79.94951630,0.01000,0.480
5,mlp.down_proj,0.71073961,0.01000,2.018
6,self_attn.k_proj,32.50390625,0.01000,0.467
6,self_attn.v_proj,3.48658395,0.01000,0.467
6,self_attn.q_proj,50.43141556,0.01000,0.477
6,self_attn.o_proj,0.32279876,0.01000,0.473
6,mlp.up_proj,43.27687836,0.01000,0.481
6,mlp.gate_proj,79.23511505,0.01000,0.478
6,mlp.down_proj,0.72877872,0.01000,2.031
7,self_attn.k_proj,33.91743851,0.01000,0.475
7,self_attn.v_proj,4.19138622,0.01000,0.469
7,self_attn.q_proj,63.09676743,0.01000,0.482
7,self_attn.o_proj,0.31201273,0.01000,0.481
7,mlp.up_proj,48.48965454,0.01000,0.484
7,mlp.gate_proj,79.57895660,0.01000,0.493
7,mlp.down_proj,0.84213424,0.01000,1.988
8,self_attn.k_proj,42.15065765,0.01000,0.471
8,self_attn.v_proj,4.18408966,0.01000,0.469
8,self_attn.q_proj,69.34562683,0.01000,0.479
8,self_attn.o_proj,0.36488774,0.01000,0.819
8,mlp.up_proj,56.41041946,0.01000,0.480
8,mlp.gate_proj,89.70135498,0.01000,0.482
8,mlp.down_proj,1.05656636,0.01000,1.997
9,self_attn.k_proj,35.16567993,0.01000,0.469
9,self_attn.v_proj,5.07817173,0.01000,0.469
9,self_attn.q_proj,86.08435059,0.01000,0.470
9,self_attn.o_proj,0.62493443,0.01000,0.472
9,mlp.up_proj,64.02466583,0.01000,0.482
9,mlp.gate_proj,104.88742828,0.01000,0.485
9,mlp.down_proj,1.33367336,0.01000,2.080
10,self_attn.k_proj,43.94557190,0.01000,0.473
10,self_attn.v_proj,5.46747303,0.01000,0.471
10,self_attn.q_proj,92.83677673,0.01000,0.472
10,self_attn.o_proj,0.59127998,0.01000,0.470
10,mlp.up_proj,79.12550354,0.01000,0.480
10,mlp.gate_proj,125.22569275,0.01000,0.480
10,mlp.down_proj,1.87667823,0.01000,2.018
11,self_attn.k_proj,51.54523468,0.01000,0.472
11,self_attn.v_proj,5.58656693,0.01000,0.466
11,self_attn.q_proj,90.90484619,0.01000,0.474
11,self_attn.o_proj,0.48921397,0.01000,0.472
11,mlp.up_proj,96.23802185,0.01000,0.478
11,mlp.gate_proj,149.39210510,0.01000,0.478
11,mlp.down_proj,2.70367146,0.01000,2.027
12,self_attn.k_proj,49.63491821,0.01000,0.471
12,self_attn.v_proj,5.90558290,0.01000,0.466
12,self_attn.q_proj,93.67311096,0.01000,0.479
12,self_attn.o_proj,0.51877552,0.01000,0.469
12,mlp.up_proj,112.33209229,0.01000,0.478
12,mlp.gate_proj,165.25785828,0.01000,0.480
12,mlp.down_proj,4.04239559,0.01000,2.022
13,self_attn.k_proj,52.68299866,0.01000,0.470
13,self_attn.v_proj,10.03530312,0.01000,0.470
13,self_attn.q_proj,107.74049377,0.01000,0.477
13,self_attn.o_proj,0.80708027,0.01000,0.469
13,mlp.up_proj,133.44732666,0.01000,0.480
13,mlp.gate_proj,179.16419983,0.01000,0.479
13,mlp.down_proj,5.91810417,0.01000,2.132
14,self_attn.k_proj,49.46713257,0.01000,0.467
14,self_attn.v_proj,18.70763016,0.01000,0.465
14,self_attn.q_proj,103.55518341,0.01000,0.468
14,self_attn.o_proj,1.71897697,0.01000,0.470
14,mlp.up_proj,148.42242432,0.01000,0.479
14,mlp.gate_proj,214.55715942,0.01000,0.480
14,mlp.down_proj,7.16841698,0.01000,2.001
15,self_attn.k_proj,46.09174347,0.01000,0.473
15,self_attn.v_proj,20.09922600,0.01000,0.469
15,self_attn.q_proj,97.27299500,0.01000,0.473
15,self_attn.o_proj,4.08784580,0.01000,0.472
15,mlp.up_proj,185.49623108,0.01000,0.479
15,mlp.gate_proj,249.19700623,0.01000,0.481
15,mlp.down_proj,15.86927128,0.01000,2.031