Commit Graph

240 Commits

Author SHA1 Message Date
Colin 2a09b9d9b1 Update meaning of stride mask label. 2025-08-11 16:07:32 +08:00
Colin 3f0eedfef8 Refine meaning dataset cache files. 2025-08-10 22:16:10 +08:00
Colin 3e52285ab6 Reset Base config. 2025-08-10 21:13:37 +08:00
Colin 133187d7bd Refine vocab_size of meaning dataset. 2025-08-10 21:01:26 +08:00
Colin a70e19df5d Init status. 2025-08-10 15:25:14 +08:00
Colin b56bbb2952 Refine import code. 2025-08-10 15:10:20 +08:00
Colin b62444a9dc Refine meaning dataset import. 2025-08-10 14:30:51 +08:00
Colin 71ab0bb57d Add config of stride with_tree end, Rename dataset to meaning. 2025-08-10 13:17:04 +08:00
Colin 1c7635556f Reconfig train model and code. 2025-08-09 16:32:48 +08:00
Colin 9c75b8920d Add stride support in meaning_dataset. 2025-08-07 17:36:43 +08:00
Colin d7191003e0 Add parent support in meaning_dataset. 2025-08-04 16:10:46 +08:00
Colin 8882073978 Update test model. 2025-08-03 18:23:26 +08:00
Colin ab65f98e39 Add weight dump. 2025-06-27 13:49:15 +08:00
Colin 9acd89c98e Update binary readme. 2025-06-27 12:23:56 +08:00
Colin a7e0bd508c Update get_relation_distance 2025-06-26 21:42:38 +08:00
Colin 927c98e823 Add meaning dataset get_relation_table support and paint to qk image. 2025-06-25 20:13:48 +08:00
Colin d8539b6b2b Update mnist to higher accuracy. 2025-06-13 20:55:43 +08:00
Colin a68e5ba5ee Update binary mnist. 2025-06-12 14:51:07 +08:00
Colin 3d2ff85fc0 Update mnist. 2025-06-10 23:14:27 +08:00
Colin 6cb969ac3b Refine LUT repeat from group to LUT. 2025-06-10 20:03:23 +08:00
Colin c322ee8228 Update Lut forward and backward. 2025-06-10 10:52:26 +08:00
Colin a676025d20 Update speed test. 2025-06-09 17:54:05 +08:00
Colin fa15680aa6 Add LNN. 2025-06-09 16:00:13 +08:00
Colin 50fb9bf6dc Add profile function. 2025-06-09 15:57:53 +08:00
Colin 878c690ac4 Refine LUTCNN, keep accuracy to ~93 2025-06-09 15:56:03 +08:00
Colin 5d03634595 Revert mnist to new backwoard of LUT 855296be55. 2025-06-09 15:04:53 +08:00
Colin 924a0ca9b4 Fix and update binary network. 2025-06-09 14:41:37 +08:00
Colin 710c901f5e Refine llamafactory train. 2025-06-08 16:01:19 +08:00
Colin 3ceb02a6b0 Update llamafactory qwen3. 2025-06-06 17:26:32 +08:00
Colin 392f507945 Add llamafactory, Refine binary LUT add SimpleLNN. 2025-06-06 17:25:04 +08:00
Colin 9a8434df61 Update demo and readme. 2025-06-05 13:17:01 +08:00
Colin eed28ac06e Add finetune/. 2025-06-01 12:31:49 +08:00
Colin e2c8668a1b Update mnist use gather to speed up. 2025-05-28 16:05:34 +08:00
Colin dd71a5dedd Refine mnist LUT by add lutCNN. 2025-05-27 18:51:07 +08:00
Colin 30dd319d8c Update mnist. 2025-05-26 18:47:55 +08:00
Colin 855296be55 set the channel to 80 at first layer accuracy to 94. 2025-05-25 15:33:39 +08:00
Colin 3eb711a97e Update wit. 2025-05-22 15:26:43 +08:00
Colin f98a951b58 Update mnist. 2025-05-22 15:23:41 +08:00
Colin f1124bc3b1 Update binary mnist. 2025-05-21 11:29:15 +08:00
Colin 9194595716 Add binary/mnist.py. 2025-05-20 14:07:10 +08:00
Colin db97131caf Update train code. 2025-04-09 19:12:21 +08:00
Colin cda7f04e49 Fix model path. 2025-03-18 15:58:08 +08:00
Colin 7faf629d45 Refine seed config. 2025-03-14 17:38:24 +08:00
Colin e3493163f3 Update configuration to str for tensorboard. 2025-03-13 23:02:11 +08:00
Colin b3817f84fe Refine model of wit. 2025-03-13 16:52:33 +08:00
Colin f411b1cc5e donot use auto optimizer. 2025-03-13 14:28:53 +08:00
Colin 990e27ba15 Fix train define. 2025-03-12 20:02:02 +08:00
Colin 90e94db2c1 Rename QwenModule to lightmodule. 2025-03-10 19:14:47 +08:00
Colin 1efda9fe25 Update rwkv train. 2025-03-10 16:26:53 +08:00
Colin 0600d46f2f Add safe softmax demo code. 2025-03-09 14:39:34 +08:00