25 days ago
-
Remove useless checkpoint folder sftkl_gxpo_k3_a0.1_b0.1_flat14k_lr1e-6_seed42_adamwdir
-
Remove useless checkpoint folder sfpo_k10_a03_w50
-
Remove useless checkpoint folder qwen25_math_7b_gxpo_adamw_k10_a03_step400
-
Remove useless checkpoint folder qwen25_math_1p5b_gxpo_muon_k10_a0.3
-
Remove useless checkpoint folder qwen25_math_1p5b_gxpo_k3_seed3407
-
Remove useless checkpoint folder qwen25_math_1p5b_gxpo_adamw_k10_a03_tput0902_resume_step400
-
Remove useless checkpoint folder qwen25_math_1p5b_gxpo_adamw_k10_a03_tput0902_best_step120
-
Remove useless checkpoint folder qwen25_math_1p5b_grpo_muon
-
Remove useless checkpoint folder llama32_3b_sft_gxpo_knights_and_knaves_fsdp2_k3_a0.8_b32_seed42
-
Remove useless checkpoint folder llama32_3b_sft_baseline_knights_and_knaves_fsdp2_b32_seed42
-
Remove useless checkpoint folder llama32_3b_grpo_b64_mb16_seed3407_v1
-
Remove useless checkpoint folder gxpo_muon_k10_a03_b64_mb16_gpu2_fsdp1_v6_memsafe
-
Remove useless checkpoint folder gxpo_k5_a05_perm_liger
-
Remove useless checkpoint folder gxpo_k3_a03_adamwdir_dynfilt
-
Remove useless checkpoint folder gxpo_k10_a05_b256_mb64_w50_tau3_v1
-
Remove useless checkpoint folder gxpo_k10_a03_w50_tau3
-
Remove useless checkpoint folder global_step_400
-
Remove useless checkpoint folder global_step_295
-
Remove useless checkpoint folder distillation
-
Upload gxpo_k3_a03_adamwdir_dynfilt/best_checkpoint (step 195, pass@1 0.3914)
-
Delete GRPO step-400 merged checkpoint
-
Upload GRPO Qwen2.5-Math-1.5B checkpoint at global step 400
-
Upload GXPO Muon step 360 actor weights
-
Upload SFT GXPO global_step_224
-
Upload complete checkpoint qwen25_math_1p5b_gxpo_muon_k10_a0.3/best_checkpoint
-
Upload complete checkpoint qwen25_math_1p5b_gxpo_muon_k10_a0.3/global_step_400
-
Upload complete checkpoint qwen25_math_1p5b_grpo_muon/best_checkpoint
-
Upload complete checkpoint qwen25_math_1p5b_grpo_muon/global_step_400
-
Upload Qwen2.5-Math-1.5B GXPO resumable checkpoint step 400
-
Upload Qwen2.5-Math-1.5B GXPO best model step 120
-
Upload Qwen 7B GXPO model-only checkpoint step 400: qwen25_math_7b_gxpo_adamw_k10_a03_step400/model_world_size_4_rank_3.pt
-
Upload Qwen 7B GXPO model-only checkpoint step 400: qwen25_math_7b_gxpo_adamw_k10_a03_step400/model_world_size_4_rank_2.pt
-
Upload Qwen 7B GXPO model-only checkpoint step 400: qwen25_math_7b_gxpo_adamw_k10_a03_step400/model_world_size_4_rank_1.pt
-
Upload Qwen 7B GXPO model-only checkpoint step 400: qwen25_math_7b_gxpo_adamw_k10_a03_step400/model_world_size_4_rank_0.pt
-
Upload Qwen 7B GXPO model-only checkpoint step 400: qwen25_math_7b_gxpo_adamw_k10_a03_step400/chat_template.jinja
-
Upload Qwen 7B GXPO model-only checkpoint step 400: qwen25_math_7b_gxpo_adamw_k10_a03_step400/tokenizer.json
-
Upload Qwen 7B GXPO model-only checkpoint step 400: qwen25_math_7b_gxpo_adamw_k10_a03_step400/tokenizer_config.json
-
Upload Qwen 7B GXPO model-only checkpoint step 400: qwen25_math_7b_gxpo_adamw_k10_a03_step400/config.json
-
Upload math distillation transactional GXPO checkpoint step 5
-
Upload math distillation AdamW checkpoint step 15
-
Upload Llama 3.2 SFT best checkpoint
-
Upload Llama 3.2 SFT best checkpoint
-
Upload recovered GXPO Muon global_step_120 weights
-
Upload GXPO Muon Qwen2.5-Math-1.5B global_step_400
-
Upload GXPO Muon Qwen2.5-Math-1.5B global_step_120
-
Upload GXPO k5 alpha0.5 Liger checkpoint step 265
-
Upload Qwen2.5-Math-1.5B GXPO checkpoint global_step_180
-
Upload llama32_3b_grpo_b64_mb16_seed3407_v1/global_step_400
-
Upload llama32_3b_grpo_b64_mb16_seed3407_v1/global_step_290
-
Upload gxpo_k10_a03_w50_tau3/global_step_400