AGILLM-4 / remote_verify_m_fold_agillm4.json
OpenTransformer's picture
Prepare AGILLM4 AR SAT NAT scaffold
528e0d6 verified
Raw
History Blame Contribute Delete
3.45 kB
[
{
"backend": "manual",
"d": 32,
"dk": 16,
"expected_k_width": 16,
"heads": 2,
"ok": true,
"preset": "pico_1x",
"rank": 16,
"rows": {
"cache_k_width": 16.0,
"cache_v_width": 16.0,
"cached_append_forward": 0.0,
"causal_alibi_forward": 0.0,
"causal_alibi_loss": 0.0,
"causal_alibi_param_grad": 0.0,
"causal_alibi_x_grad": 0.0,
"none_forward": 0.0,
"none_loss": 0.0,
"none_param_grad": 0.0,
"none_x_grad": 0.0,
"sat_alibi_forward": 0.0,
"sat_alibi_loss": 0.0,
"sat_alibi_param_grad": 0.0,
"sat_alibi_x_grad": 0.0
},
"tol": 0.0002
},
{
"backend": "sdpa",
"d": 32,
"dk": 16,
"expected_k_width": 16,
"heads": 2,
"ok": true,
"preset": "pico_1x",
"rank": 16,
"rows": {
"cache_k_width": 16.0,
"cache_v_width": 16.0,
"cached_append_forward": 0.0,
"causal_alibi_forward": 5.386769771575928e-06,
"causal_alibi_loss": 3.725290298461914e-09,
"causal_alibi_param_grad": 5.189329385757446e-06,
"causal_alibi_x_grad": 8.444767445325851e-07,
"none_forward": 1.4901161193847656e-08,
"none_loss": 9.313225746154785e-10,
"none_param_grad": 2.0703300833702087e-06,
"none_x_grad": 1.7954153008759022e-07,
"sat_alibi_forward": 0.0,
"sat_alibi_loss": 0.0,
"sat_alibi_param_grad": 8.198898285627365e-06,
"sat_alibi_x_grad": 9.0012326836586e-07
},
"tol": 0.0002
},
{
"backend": "manual",
"d": 128,
"dk": 16,
"expected_k_width": 16,
"heads": 8,
"ok": true,
"preset": "micro_3x",
"rank": 48,
"rows": {
"cache_k_width": 16.0,
"cache_v_width": 16.0,
"cached_append_forward": 0.0,
"causal_alibi_forward": 0.00011304020881652832,
"causal_alibi_loss": 1.043081283569336e-07,
"causal_alibi_param_grad": 1.3775425031781197e-06,
"causal_alibi_x_grad": 1.592270564287901e-07,
"metric_cache_cleared_on_train": 0.0,
"metric_cache_reused": 0.0,
"none_forward": 8.377432823181152e-05,
"none_loss": 1.30385160446167e-08,
"none_param_grad": 5.532929208129644e-07,
"none_x_grad": 3.541026671882719e-08,
"sat_alibi_forward": 0.00014029256999492645,
"sat_alibi_loss": 9.12696123123169e-08,
"sat_alibi_param_grad": 9.688083082437515e-07,
"sat_alibi_x_grad": 1.5371188055723906e-07
},
"tol": 0.0002
},
{
"backend": "sdpa",
"d": 128,
"dk": 16,
"expected_k_width": 16,
"heads": 8,
"ok": true,
"preset": "micro_3x",
"rank": 48,
"rows": {
"cache_k_width": 16.0,
"cache_v_width": 16.0,
"cached_append_forward": 0.0,
"causal_alibi_forward": 0.00011304020881652832,
"causal_alibi_loss": 1.0244548320770264e-07,
"causal_alibi_param_grad": 1.3635726645588875e-06,
"causal_alibi_x_grad": 1.8296123016625643e-07,
"metric_cache_cleared_on_train": 0.0,
"metric_cache_reused": 0.0,
"none_forward": 8.377432823181152e-05,
"none_loss": 5.587935447692871e-09,
"none_param_grad": 5.358306225389242e-07,
"none_x_grad": 3.1091985874809325e-08,
"sat_alibi_forward": 0.00014029256999492645,
"sat_alibi_loss": 9.12696123123169e-08,
"sat_alibi_param_grad": 1.071603037416935e-06,
"sat_alibi_x_grad": 1.1662632459774613e-07
},
"tol": 0.0002
}
]