diff --git a/.gitattributes b/.gitattributes index 94bc834eeee8852f65a7648efb06c2ecd502ba38..caa358f08503468f2415ec1949ad197b44df80d5 100644 --- a/.gitattributes +++ b/.gitattributes @@ -45,3 +45,11 @@ mathmedium2-mutualposclip0.5-llama3.2-3b-it-oldreward-4k-corrupt-0.8/global_step mathmedium2-mutualposclip0.5-llama3.2-3b-it-oldreward-4k-corrupt-0.8/global_step_100/actor/huggingface/tokenizer.json filter=lfs diff=lfs merge=lfs -text mathmedium2-mutualposclip0.5-llama3.2-3b-it-oldreward-4k-corrupt-0.8/global_step_200/actor/huggingface/tokenizer.json filter=lfs diff=lfs merge=lfs -text mathmedium2-mutualposclip0.5-qwen2.5-math-1.5b-oldreward-corrupt_delete-0.8/global_step_100/actor/global_step_100/tokenizer.json filter=lfs diff=lfs merge=lfs -text +mathmedium2-mutualposclip0.5-qwen2.5-math-1.5b-oldreward-corrupt_delete-0.8/global_step_200/actor/global_step_200/tokenizer.json filter=lfs diff=lfs merge=lfs -text +mathmedium2-mutualposclip-0.0-llama3.2-3b-it-oldreward-4k-corrupt-0.8/global_step_100/actor/huggingface/tokenizer.json filter=lfs diff=lfs merge=lfs -text +mathmedium2-mutualposclip-0.0-llama3.2-3b-it-oldreward-4k-corrupt-0.8/global_step_200/actor/huggingface/tokenizer.json filter=lfs diff=lfs merge=lfs -text +pilecc_cont1-r1-ppo-llama3.2-3b-it-em-warmup-0.05-rouge-rougeL-temperature-1.0/critic/global_step_300/tokenizer.json filter=lfs diff=lfs merge=lfs -text +pilecc_cont1-r1-ppo-llama3.2-3b-it-em-warmup-0.05-rouge-rougeL-temperature-1.0/actor/global_step_300/tokenizer.json filter=lfs diff=lfs merge=lfs -text +pilecc_cont2-r1-ppo-llama3.2-3b-it-em-warmup-0.05-rouge-rougeL-temperature-1.0/critic/global_step_300/tokenizer.json filter=lfs diff=lfs merge=lfs -text +pilecc_cont2-r1-ppo-llama3.2-3b-it-em-warmup-0.05-rouge-rougeL-temperature-1.0/actor/global_step_300/tokenizer.json filter=lfs diff=lfs merge=lfs -text +mathhard2-grpo-qwen3-1.7b-4k-step-100/tokenizer.json filter=lfs diff=lfs merge=lfs -text diff --git a/mathhard2-grpo-qwen3-1.7b-4k-step-100/model.safetensors b/mathhard2-grpo-qwen3-1.7b-4k-step-100/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..fafec799b41fcbddbc62e145de3efad662d6bd10 --- /dev/null +++ b/mathhard2-grpo-qwen3-1.7b-4k-step-100/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:91010ca819070628c7bc2f2f608035f1501ba529461bce09f4582cdb9db79320 +size 4063515640 diff --git a/mathhard2-grpo-qwen3-1.7b-4k-step-100/tokenizer.json b/mathhard2-grpo-qwen3-1.7b-4k-step-100/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..cd71f61a15a522601badb3dc960d800d9cb3766c --- /dev/null +++ b/mathhard2-grpo-qwen3-1.7b-4k-step-100/tokenizer.json @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:aeb13307a71acd8fe81861d94ad54ab689df773318809eed3cbe794b4492dae4 +size 11422654 diff --git a/mathmedium2-mutualposclip-0.0-llama3.2-3b-it-oldreward-4k-corrupt-0.8/global_step_100/actor/extra_state_world_size_8_rank_2.pt b/mathmedium2-mutualposclip-0.0-llama3.2-3b-it-oldreward-4k-corrupt-0.8/global_step_100/actor/extra_state_world_size_8_rank_2.pt new file mode 100644 index 0000000000000000000000000000000000000000..39456853b19bbb092a4a4bbe4f75b8d5aed57b94 --- /dev/null +++ b/mathmedium2-mutualposclip-0.0-llama3.2-3b-it-oldreward-4k-corrupt-0.8/global_step_100/actor/extra_state_world_size_8_rank_2.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:992867e9821a253236805cfdcfb0704e579adc0703438bc32a8013b9e3435bc6 +size 14632 diff --git a/mathmedium2-mutualposclip-0.0-llama3.2-3b-it-oldreward-4k-corrupt-0.8/global_step_100/actor/extra_state_world_size_8_rank_5.pt b/mathmedium2-mutualposclip-0.0-llama3.2-3b-it-oldreward-4k-corrupt-0.8/global_step_100/actor/extra_state_world_size_8_rank_5.pt new file mode 100644 index 0000000000000000000000000000000000000000..db53f2926b13e92e99ba9aa945d80d17849f8768 --- /dev/null +++ b/mathmedium2-mutualposclip-0.0-llama3.2-3b-it-oldreward-4k-corrupt-0.8/global_step_100/actor/extra_state_world_size_8_rank_5.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9bb138fd25171ad6d128037d07be43c57f8165e36cb6140da796e3d707705e5e +size 14632 diff --git a/mathmedium2-mutualposclip-0.0-llama3.2-3b-it-oldreward-4k-corrupt-0.8/global_step_100/actor/huggingface/model-00001-of-00003.safetensors b/mathmedium2-mutualposclip-0.0-llama3.2-3b-it-oldreward-4k-corrupt-0.8/global_step_100/actor/huggingface/model-00001-of-00003.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..b798774dfebf5c08574444e3b7aee0865f05de2e --- /dev/null +++ b/mathmedium2-mutualposclip-0.0-llama3.2-3b-it-oldreward-4k-corrupt-0.8/global_step_100/actor/huggingface/model-00001-of-00003.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:743148d0a518bcbc54292c698e5f9e4ce8027722e14cf452471cb33bd79d49a4 +size 4998767360 diff --git a/mathmedium2-mutualposclip-0.0-llama3.2-3b-it-oldreward-4k-corrupt-0.8/global_step_100/actor/huggingface/model-00002-of-00003.safetensors b/mathmedium2-mutualposclip-0.0-llama3.2-3b-it-oldreward-4k-corrupt-0.8/global_step_100/actor/huggingface/model-00002-of-00003.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..6d0f763b2aa85aeedc0ae5cebbf1b33b3962ee02 --- /dev/null +++ b/mathmedium2-mutualposclip-0.0-llama3.2-3b-it-oldreward-4k-corrupt-0.8/global_step_100/actor/huggingface/model-00002-of-00003.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bf908dacf6705d257d03b1f61bb8bb0e09baf99f01d1443dd63484892004e129 +size 4932808968 diff --git a/mathmedium2-mutualposclip-0.0-llama3.2-3b-it-oldreward-4k-corrupt-0.8/global_step_100/actor/huggingface/model-00003-of-00003.safetensors b/mathmedium2-mutualposclip-0.0-llama3.2-3b-it-oldreward-4k-corrupt-0.8/global_step_100/actor/huggingface/model-00003-of-00003.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..3578a7c8e69541c8eb43ff12ac919651869cf511 --- /dev/null +++ b/mathmedium2-mutualposclip-0.0-llama3.2-3b-it-oldreward-4k-corrupt-0.8/global_step_100/actor/huggingface/model-00003-of-00003.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6ed581ad516b279565981378d7ee7ee7b79171c1d5f9759a5674cba7ebb083c3 +size 4495462048 diff --git a/mathmedium2-mutualposclip-0.0-llama3.2-3b-it-oldreward-4k-corrupt-0.8/global_step_100/actor/huggingface/tokenizer.json b/mathmedium2-mutualposclip-0.0-llama3.2-3b-it-oldreward-4k-corrupt-0.8/global_step_100/actor/huggingface/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..1c1d8d5c9024994f1d3b00f9662b8dd89ca13cf2 --- /dev/null +++ b/mathmedium2-mutualposclip-0.0-llama3.2-3b-it-oldreward-4k-corrupt-0.8/global_step_100/actor/huggingface/tokenizer.json @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6b9e4e7fb171f92fd137b777cc2714bf87d11576700a1dcd7a399e7bbe39537b +size 17209920 diff --git a/mathmedium2-mutualposclip-0.0-llama3.2-3b-it-oldreward-4k-corrupt-0.8/global_step_100/actor/model_world_size_8_rank_0.pt b/mathmedium2-mutualposclip-0.0-llama3.2-3b-it-oldreward-4k-corrupt-0.8/global_step_100/actor/model_world_size_8_rank_0.pt new file mode 100644 index 0000000000000000000000000000000000000000..91a009eb15fa2b3033fdad8ccf0f20d5873bd651 --- /dev/null +++ b/mathmedium2-mutualposclip-0.0-llama3.2-3b-it-oldreward-4k-corrupt-0.8/global_step_100/actor/model_world_size_8_rank_0.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3259d7f911b83c67a5275b68f453a78c1d2212c56083817e4527983cce67a853 +size 1803528186 diff --git a/mathmedium2-mutualposclip-0.0-llama3.2-3b-it-oldreward-4k-corrupt-0.8/global_step_100/actor/model_world_size_8_rank_5.pt b/mathmedium2-mutualposclip-0.0-llama3.2-3b-it-oldreward-4k-corrupt-0.8/global_step_100/actor/model_world_size_8_rank_5.pt new file mode 100644 index 0000000000000000000000000000000000000000..3114ba020e01657a00daff8ac13052a821c68bb0 --- /dev/null +++ b/mathmedium2-mutualposclip-0.0-llama3.2-3b-it-oldreward-4k-corrupt-0.8/global_step_100/actor/model_world_size_8_rank_5.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:980857611402274d9bae2c62c166224fd7c51a7a06e01de16e06af1d7554ee94 +size 1803528186 diff --git a/mathmedium2-mutualposclip-0.0-llama3.2-3b-it-oldreward-4k-corrupt-0.8/global_step_100/actor/model_world_size_8_rank_7.pt b/mathmedium2-mutualposclip-0.0-llama3.2-3b-it-oldreward-4k-corrupt-0.8/global_step_100/actor/model_world_size_8_rank_7.pt new file mode 100644 index 0000000000000000000000000000000000000000..ad7279359377284e7bb6427418eaa0cce8786e58 --- /dev/null +++ b/mathmedium2-mutualposclip-0.0-llama3.2-3b-it-oldreward-4k-corrupt-0.8/global_step_100/actor/model_world_size_8_rank_7.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f9bed3a6ee6c40fae451cd3c7824447748646191950873f1597a1dfb080100b1 +size 1803528186 diff --git a/mathmedium2-mutualposclip-0.0-llama3.2-3b-it-oldreward-4k-corrupt-0.8/global_step_200/actor/huggingface/model-00001-of-00003.safetensors b/mathmedium2-mutualposclip-0.0-llama3.2-3b-it-oldreward-4k-corrupt-0.8/global_step_200/actor/huggingface/model-00001-of-00003.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..0b089142c84cc35c53e6c81e6972e91d888e6291 --- /dev/null +++ b/mathmedium2-mutualposclip-0.0-llama3.2-3b-it-oldreward-4k-corrupt-0.8/global_step_200/actor/huggingface/model-00001-of-00003.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c3747dc4d48f10d3247cd1c6d185bdf9c6d5fbaeaa482502a55a7fc20bc02665 +size 4998767360 diff --git a/mathmedium2-mutualposclip-0.0-llama3.2-3b-it-oldreward-4k-corrupt-0.8/global_step_200/actor/huggingface/model-00002-of-00003.safetensors b/mathmedium2-mutualposclip-0.0-llama3.2-3b-it-oldreward-4k-corrupt-0.8/global_step_200/actor/huggingface/model-00002-of-00003.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..9b0885dd0dd30751848456b5b880a4a0e83b1dc9 --- /dev/null +++ b/mathmedium2-mutualposclip-0.0-llama3.2-3b-it-oldreward-4k-corrupt-0.8/global_step_200/actor/huggingface/model-00002-of-00003.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:66266fe16262dbbded13f1aeec593f266fd0033cdc919454eb05a155700c29ba +size 4932808968 diff --git a/mathmedium2-mutualposclip-0.0-llama3.2-3b-it-oldreward-4k-corrupt-0.8/global_step_200/actor/huggingface/model-00003-of-00003.safetensors b/mathmedium2-mutualposclip-0.0-llama3.2-3b-it-oldreward-4k-corrupt-0.8/global_step_200/actor/huggingface/model-00003-of-00003.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..2155be0e8496a2ec1e7010b4e3fa51dd855ce94a --- /dev/null +++ b/mathmedium2-mutualposclip-0.0-llama3.2-3b-it-oldreward-4k-corrupt-0.8/global_step_200/actor/huggingface/model-00003-of-00003.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:da5dffd3991c8f414d8afe3e447599121865b7496f31a461d7a5393ca9525b97 +size 4495462048 diff --git a/mathmedium2-mutualposclip-0.0-llama3.2-3b-it-oldreward-4k-corrupt-0.8/global_step_200/actor/huggingface/tokenizer.json b/mathmedium2-mutualposclip-0.0-llama3.2-3b-it-oldreward-4k-corrupt-0.8/global_step_200/actor/huggingface/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..1c1d8d5c9024994f1d3b00f9662b8dd89ca13cf2 --- /dev/null +++ b/mathmedium2-mutualposclip-0.0-llama3.2-3b-it-oldreward-4k-corrupt-0.8/global_step_200/actor/huggingface/tokenizer.json @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6b9e4e7fb171f92fd137b777cc2714bf87d11576700a1dcd7a399e7bbe39537b +size 17209920 diff --git a/mathmedium2-mutualposclip-0.0-llama3.2-3b-it-oldreward-4k-corrupt-0.8/global_step_200/data.pt b/mathmedium2-mutualposclip-0.0-llama3.2-3b-it-oldreward-4k-corrupt-0.8/global_step_200/data.pt new file mode 100644 index 0000000000000000000000000000000000000000..8d43c8725f44995cf5ecb67f3eab528ce342d4cc --- /dev/null +++ b/mathmedium2-mutualposclip-0.0-llama3.2-3b-it-oldreward-4k-corrupt-0.8/global_step_200/data.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:80905c80177b8dac7924e4499bb593196cba24fee0013f6f28089c0e28a55eb9 +size 1492 diff --git a/mathmedium2-mutualposclip0.5-llama3.2-3b-it-oldreward-4k-corrupt-0.2/global_step_300/actor/extra_state_world_size_8_rank_1.pt b/mathmedium2-mutualposclip0.5-llama3.2-3b-it-oldreward-4k-corrupt-0.2/global_step_300/actor/extra_state_world_size_8_rank_1.pt new file mode 100644 index 0000000000000000000000000000000000000000..d198d0e36fd5075f9971936324656e05a6f4c5ca --- /dev/null +++ b/mathmedium2-mutualposclip0.5-llama3.2-3b-it-oldreward-4k-corrupt-0.2/global_step_300/actor/extra_state_world_size_8_rank_1.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e4a434f32f69e38311bd68c37d840f6734f925181dab3e88e786e79af4c36f1e +size 14760 diff --git a/mathmedium2-mutualposclip0.5-llama3.2-3b-it-oldreward-4k-corrupt-0.2/global_step_300/actor/extra_state_world_size_8_rank_3.pt b/mathmedium2-mutualposclip0.5-llama3.2-3b-it-oldreward-4k-corrupt-0.2/global_step_300/actor/extra_state_world_size_8_rank_3.pt new file mode 100644 index 0000000000000000000000000000000000000000..9bfab2bf2399c28a7a734c19b0e38284563fd0c8 --- /dev/null +++ b/mathmedium2-mutualposclip0.5-llama3.2-3b-it-oldreward-4k-corrupt-0.2/global_step_300/actor/extra_state_world_size_8_rank_3.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9ce2963ff99c06d4610f89af7431bc4c53bc8d2a9af981561f90a854796bbb19 +size 14696 diff --git a/mathmedium2-mutualposclip0.5-llama3.2-3b-it-oldreward-4k-corrupt-0.2/global_step_300/actor/extra_state_world_size_8_rank_4.pt b/mathmedium2-mutualposclip0.5-llama3.2-3b-it-oldreward-4k-corrupt-0.2/global_step_300/actor/extra_state_world_size_8_rank_4.pt new file mode 100644 index 0000000000000000000000000000000000000000..9c4f01cd7f328790de533ec7f38be986d3d9fab6 --- /dev/null +++ b/mathmedium2-mutualposclip0.5-llama3.2-3b-it-oldreward-4k-corrupt-0.2/global_step_300/actor/extra_state_world_size_8_rank_4.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3b601ee254c3732d86b5e9ad5e5fb165927c31d0b1fcaf2d6f5e7ec625c23b53 +size 14696 diff --git a/mathmedium2-mutualposclip0.5-llama3.2-3b-it-oldreward-4k-corrupt-0.2/global_step_300/actor/extra_state_world_size_8_rank_6.pt b/mathmedium2-mutualposclip0.5-llama3.2-3b-it-oldreward-4k-corrupt-0.2/global_step_300/actor/extra_state_world_size_8_rank_6.pt new file mode 100644 index 0000000000000000000000000000000000000000..4f5589b2efa67800484655b7da796d9a1293e42f --- /dev/null +++ b/mathmedium2-mutualposclip0.5-llama3.2-3b-it-oldreward-4k-corrupt-0.2/global_step_300/actor/extra_state_world_size_8_rank_6.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a62afb11bc86fce9c8b0d999d970bb9c6298fb1d33f0752a581f25d825deec65 +size 14760 diff --git a/mathmedium2-mutualposclip0.5-llama3.2-3b-it-oldreward-4k-corrupt-0.2/global_step_300/actor/model_world_size_8_rank_1.pt b/mathmedium2-mutualposclip0.5-llama3.2-3b-it-oldreward-4k-corrupt-0.2/global_step_300/actor/model_world_size_8_rank_1.pt new file mode 100644 index 0000000000000000000000000000000000000000..e0db4744bda05f22b2996ded49c2f9a78d414f1e --- /dev/null +++ b/mathmedium2-mutualposclip0.5-llama3.2-3b-it-oldreward-4k-corrupt-0.2/global_step_300/actor/model_world_size_8_rank_1.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8e03e4be48abde3073aa84752897980cf036f8204818ce9077105b863f3fc151 +size 1803528186 diff --git a/mathmedium2-mutualposclip0.5-llama3.2-3b-it-oldreward-4k-corrupt-0.2/global_step_300/actor/model_world_size_8_rank_3.pt b/mathmedium2-mutualposclip0.5-llama3.2-3b-it-oldreward-4k-corrupt-0.2/global_step_300/actor/model_world_size_8_rank_3.pt new file mode 100644 index 0000000000000000000000000000000000000000..f2ece43267f5ad3a6ab2c55be2e14461b533c0ba --- /dev/null +++ b/mathmedium2-mutualposclip0.5-llama3.2-3b-it-oldreward-4k-corrupt-0.2/global_step_300/actor/model_world_size_8_rank_3.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:64c0ccd4fb21c80cae54d53b08b7f2ea9e8f60b8f5d87f0cd5c2f69119893f29 +size 1803528186 diff --git a/mathmedium2-mutualposclip0.5-llama3.2-3b-it-oldreward-4k-corrupt-0.2/global_step_300/actor/model_world_size_8_rank_4.pt b/mathmedium2-mutualposclip0.5-llama3.2-3b-it-oldreward-4k-corrupt-0.2/global_step_300/actor/model_world_size_8_rank_4.pt new file mode 100644 index 0000000000000000000000000000000000000000..4858daedb9b03ec2e3b81c071c697adbcd3db470 --- /dev/null +++ b/mathmedium2-mutualposclip0.5-llama3.2-3b-it-oldreward-4k-corrupt-0.2/global_step_300/actor/model_world_size_8_rank_4.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:66aabf9822550b7861dffdfaf0ded777b4a067b24a263b8207a83eba48cac63a +size 1803528186 diff --git a/mathmedium2-mutualposclip0.5-llama3.2-3b-it-oldreward-4k-corrupt-0.2/global_step_300/actor/model_world_size_8_rank_6.pt b/mathmedium2-mutualposclip0.5-llama3.2-3b-it-oldreward-4k-corrupt-0.2/global_step_300/actor/model_world_size_8_rank_6.pt new file mode 100644 index 0000000000000000000000000000000000000000..d545fb8d4b2174341c1ea52c5fbe6fed6b3592c9 --- /dev/null +++ b/mathmedium2-mutualposclip0.5-llama3.2-3b-it-oldreward-4k-corrupt-0.2/global_step_300/actor/model_world_size_8_rank_6.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:584e43bb61bd591ad02a47a4bf21f634216a0d0532634ce8853bdca9060c59b0 +size 1803528186 diff --git a/mathmedium2-mutualposclip0.5-llama3.2-3b-it-oldreward-4k-corrupt-0.2/global_step_300/actor/optim_world_size_8_rank_1.pt b/mathmedium2-mutualposclip0.5-llama3.2-3b-it-oldreward-4k-corrupt-0.2/global_step_300/actor/optim_world_size_8_rank_1.pt new file mode 100644 index 0000000000000000000000000000000000000000..77428da6042f26d4c40eaf44134400947938ae29 --- /dev/null +++ b/mathmedium2-mutualposclip0.5-llama3.2-3b-it-oldreward-4k-corrupt-0.2/global_step_300/actor/optim_world_size_8_rank_1.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:826173794c1462c11c88eb4a9ce817dfd1546797d5940c2a6b34159d1f96e1d0 +size 3212775159 diff --git a/mathmedium2-mutualposclip0.5-llama3.2-3b-it-oldreward-4k-corrupt-0.2/global_step_300/actor/optim_world_size_8_rank_3.pt b/mathmedium2-mutualposclip0.5-llama3.2-3b-it-oldreward-4k-corrupt-0.2/global_step_300/actor/optim_world_size_8_rank_3.pt new file mode 100644 index 0000000000000000000000000000000000000000..11fdf7d0e81cd6b2a7a456dfa640c486c49e9fa0 --- /dev/null +++ b/mathmedium2-mutualposclip0.5-llama3.2-3b-it-oldreward-4k-corrupt-0.2/global_step_300/actor/optim_world_size_8_rank_3.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fb2727ab149b180075f74af075723ec2d0b465c9bfa8a798f80fdb992bb99037 +size 3212775159 diff --git a/mathmedium2-mutualposclip0.5-llama3.2-3b-it-oldreward-4k-corrupt-0.2/global_step_300/actor/optim_world_size_8_rank_4.pt b/mathmedium2-mutualposclip0.5-llama3.2-3b-it-oldreward-4k-corrupt-0.2/global_step_300/actor/optim_world_size_8_rank_4.pt new file mode 100644 index 0000000000000000000000000000000000000000..6fb7a64dee05748bc061840ff4c7813e389e90ad --- /dev/null +++ b/mathmedium2-mutualposclip0.5-llama3.2-3b-it-oldreward-4k-corrupt-0.2/global_step_300/actor/optim_world_size_8_rank_4.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:98162afec6722951303c3fc1ce68eea96e7299372449ee6338360c7d2ce87a0d +size 3212775159 diff --git a/mathmedium2-mutualposclip0.5-llama3.2-3b-it-oldreward-4k-corrupt-0.2/global_step_300/actor/optim_world_size_8_rank_6.pt b/mathmedium2-mutualposclip0.5-llama3.2-3b-it-oldreward-4k-corrupt-0.2/global_step_300/actor/optim_world_size_8_rank_6.pt new file mode 100644 index 0000000000000000000000000000000000000000..27bb1f6f3c21b649c9222deeb5f2b11570cd0666 --- /dev/null +++ b/mathmedium2-mutualposclip0.5-llama3.2-3b-it-oldreward-4k-corrupt-0.2/global_step_300/actor/optim_world_size_8_rank_6.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f0a2efc1e702d818f1380f672012ef1fd4b99e5515814b709241f20b0659dc9c +size 3212775159 diff --git a/mathmedium2-mutualposclip0.5-llama3.2-3b-it-oldreward-4k-corrupt-0.2/global_step_300/data.pt b/mathmedium2-mutualposclip0.5-llama3.2-3b-it-oldreward-4k-corrupt-0.2/global_step_300/data.pt new file mode 100644 index 0000000000000000000000000000000000000000..73d3d2e83b5b4d285bf5dd14c48f23f12794854b --- /dev/null +++ b/mathmedium2-mutualposclip0.5-llama3.2-3b-it-oldreward-4k-corrupt-0.2/global_step_300/data.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0a9a5343c05377b06534adffd6adb01f34763c07ccf76276b2090629a5523063 +size 1492 diff --git a/mathmedium2-mutualposclip0.5-qwen2.5-math-1.5b-oldreward-corrupt_delete-0.8/global_step_100/actor/global_step_100/model-00002-of-00002.safetensors b/mathmedium2-mutualposclip0.5-qwen2.5-math-1.5b-oldreward-corrupt_delete-0.8/global_step_100/actor/global_step_100/model-00002-of-00002.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..f235fcf64256fa832323a0130b397d85f12e7ecc --- /dev/null +++ b/mathmedium2-mutualposclip0.5-qwen2.5-math-1.5b-oldreward-corrupt_delete-0.8/global_step_100/actor/global_step_100/model-00002-of-00002.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:086158b5538eca111e7bfb6200f1a9692229371ce559c991cb2515ba53b748db +size 2111719976 diff --git a/mathmedium2-mutualposclip0.5-qwen2.5-math-1.5b-oldreward-corrupt_delete-0.8/global_step_200/actor/global_step_200/model-00001-of-00002.safetensors b/mathmedium2-mutualposclip0.5-qwen2.5-math-1.5b-oldreward-corrupt_delete-0.8/global_step_200/actor/global_step_200/model-00001-of-00002.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..5edfd38592c4384dbce9d7650417df968316713a --- /dev/null +++ b/mathmedium2-mutualposclip0.5-qwen2.5-math-1.5b-oldreward-corrupt_delete-0.8/global_step_200/actor/global_step_200/model-00001-of-00002.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:90cd2b52e6af87e7c21f5e3cd1b04850eaf362a4424530444cfde9de231cf6ca +size 4996670464 diff --git a/mathmedium2-mutualposclip0.5-qwen2.5-math-1.5b-oldreward-corrupt_delete-0.8/global_step_200/actor/global_step_200/model-00002-of-00002.safetensors b/mathmedium2-mutualposclip0.5-qwen2.5-math-1.5b-oldreward-corrupt_delete-0.8/global_step_200/actor/global_step_200/model-00002-of-00002.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..11d0e40f0569d071f7bd2a3996c8a1f63f0ff5da --- /dev/null +++ b/mathmedium2-mutualposclip0.5-qwen2.5-math-1.5b-oldreward-corrupt_delete-0.8/global_step_200/actor/global_step_200/model-00002-of-00002.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c3ef2a764aaf15213949fa3000966dc7fcf75b7c4481c75791e2527f974f0b50 +size 2111719976 diff --git a/mathmedium2-mutualposclip0.5-qwen2.5-math-1.5b-oldreward-corrupt_delete-0.8/global_step_200/actor/global_step_200/tokenizer.json b/mathmedium2-mutualposclip0.5-qwen2.5-math-1.5b-oldreward-corrupt_delete-0.8/global_step_200/actor/global_step_200/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..51ebb3ba93988c73bbd83a3c163a14343fbc5b06 --- /dev/null +++ b/mathmedium2-mutualposclip0.5-qwen2.5-math-1.5b-oldreward-corrupt_delete-0.8/global_step_200/actor/global_step_200/tokenizer.json @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9c5ae00e602b8860cbd784ba82a8aa14e8feecec692e7076590d014d7b7fdafa +size 11421896 diff --git a/mathmedium2-mutualposclip0.5-qwen2.5-math-1.5b-oldreward-corrupt_delete-0.8/global_step_200/random_state.pkl b/mathmedium2-mutualposclip0.5-qwen2.5-math-1.5b-oldreward-corrupt_delete-0.8/global_step_200/random_state.pkl new file mode 100644 index 0000000000000000000000000000000000000000..19c684a2da262da95e3759b7391add3ddf7c5a1a --- /dev/null +++ b/mathmedium2-mutualposclip0.5-qwen2.5-math-1.5b-oldreward-corrupt_delete-0.8/global_step_200/random_state.pkl @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:dfa16f5bd60a1e1c60b5150671b96e30d4d670310cbe4a6e5eb00f2b0b387205 +size 12846 diff --git a/pilecc_cont1-r1-ppo-llama3.2-3b-it-em-warmup-0.05-rouge-rougeL-temperature-1.0/actor/global_step_300/model-00001-of-00003.safetensors b/pilecc_cont1-r1-ppo-llama3.2-3b-it-em-warmup-0.05-rouge-rougeL-temperature-1.0/actor/global_step_300/model-00001-of-00003.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..9e8b57f70a8bc35e8e163feeb739ffd476ac371a --- /dev/null +++ b/pilecc_cont1-r1-ppo-llama3.2-3b-it-em-warmup-0.05-rouge-rougeL-temperature-1.0/actor/global_step_300/model-00001-of-00003.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0e556a6861d3168ce979ea9ef70c191eb01991b4a3b17156c009b580a32db176 +size 4998767360 diff --git a/pilecc_cont1-r1-ppo-llama3.2-3b-it-em-warmup-0.05-rouge-rougeL-temperature-1.0/actor/global_step_300/model-00002-of-00003.safetensors b/pilecc_cont1-r1-ppo-llama3.2-3b-it-em-warmup-0.05-rouge-rougeL-temperature-1.0/actor/global_step_300/model-00002-of-00003.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..3e7ee63809de9d2d44f2b0b17f2ff5dac88b9679 --- /dev/null +++ b/pilecc_cont1-r1-ppo-llama3.2-3b-it-em-warmup-0.05-rouge-rougeL-temperature-1.0/actor/global_step_300/model-00002-of-00003.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9ec715f9b572ee7e4978bd7f4531f441f671fbfcf6890975a658a68f73b30b65 +size 4932808968 diff --git a/pilecc_cont1-r1-ppo-llama3.2-3b-it-em-warmup-0.05-rouge-rougeL-temperature-1.0/actor/global_step_300/model-00003-of-00003.safetensors b/pilecc_cont1-r1-ppo-llama3.2-3b-it-em-warmup-0.05-rouge-rougeL-temperature-1.0/actor/global_step_300/model-00003-of-00003.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..0ab7ef7b32cc7416f07e79139ef87bd8dadd4cac --- /dev/null +++ b/pilecc_cont1-r1-ppo-llama3.2-3b-it-em-warmup-0.05-rouge-rougeL-temperature-1.0/actor/global_step_300/model-00003-of-00003.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:48272d722876908736addd29a1a62b381edd76eece05e236f0dbb3018e678234 +size 4495462048 diff --git a/pilecc_cont1-r1-ppo-llama3.2-3b-it-em-warmup-0.05-rouge-rougeL-temperature-1.0/actor/global_step_300/tokenizer.json b/pilecc_cont1-r1-ppo-llama3.2-3b-it-em-warmup-0.05-rouge-rougeL-temperature-1.0/actor/global_step_300/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..1c1d8d5c9024994f1d3b00f9662b8dd89ca13cf2 --- /dev/null +++ b/pilecc_cont1-r1-ppo-llama3.2-3b-it-em-warmup-0.05-rouge-rougeL-temperature-1.0/actor/global_step_300/tokenizer.json @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6b9e4e7fb171f92fd137b777cc2714bf87d11576700a1dcd7a399e7bbe39537b +size 17209920 diff --git a/pilecc_cont1-r1-ppo-llama3.2-3b-it-em-warmup-0.05-rouge-rougeL-temperature-1.0/critic/global_step_300/model-00001-of-00003.safetensors b/pilecc_cont1-r1-ppo-llama3.2-3b-it-em-warmup-0.05-rouge-rougeL-temperature-1.0/critic/global_step_300/model-00001-of-00003.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..585c7b90cd0b5c9fd3932b9fb3cd61afb744bedc --- /dev/null +++ b/pilecc_cont1-r1-ppo-llama3.2-3b-it-em-warmup-0.05-rouge-rougeL-temperature-1.0/critic/global_step_300/model-00001-of-00003.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fdd104338c78196f3f1d5ad85ba4b88cefc0b80e18a12b24ab3d910a5c93775a +size 4998767360 diff --git a/pilecc_cont1-r1-ppo-llama3.2-3b-it-em-warmup-0.05-rouge-rougeL-temperature-1.0/critic/global_step_300/model-00002-of-00003.safetensors b/pilecc_cont1-r1-ppo-llama3.2-3b-it-em-warmup-0.05-rouge-rougeL-temperature-1.0/critic/global_step_300/model-00002-of-00003.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..6240b92a7552df3585b105567eb520a0397f2a71 --- /dev/null +++ b/pilecc_cont1-r1-ppo-llama3.2-3b-it-em-warmup-0.05-rouge-rougeL-temperature-1.0/critic/global_step_300/model-00002-of-00003.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7b7e7275609977c9a1ff8fd7b96044bce00d27600a7fc855be768a5258cae819 +size 4932808968 diff --git a/pilecc_cont1-r1-ppo-llama3.2-3b-it-em-warmup-0.05-rouge-rougeL-temperature-1.0/critic/global_step_300/model-00003-of-00003.safetensors b/pilecc_cont1-r1-ppo-llama3.2-3b-it-em-warmup-0.05-rouge-rougeL-temperature-1.0/critic/global_step_300/model-00003-of-00003.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..92e4f584146a05d9a6fa34b369423ff54fc9a311 --- /dev/null +++ b/pilecc_cont1-r1-ppo-llama3.2-3b-it-em-warmup-0.05-rouge-rougeL-temperature-1.0/critic/global_step_300/model-00003-of-00003.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c3a8ce7d103edc3c38337ac87a63afb8b3824d74833ee97a7a304a7a250b82c4 +size 2919464628 diff --git a/pilecc_cont1-r1-ppo-llama3.2-3b-it-em-warmup-0.05-rouge-rougeL-temperature-1.0/critic/global_step_300/tokenizer.json b/pilecc_cont1-r1-ppo-llama3.2-3b-it-em-warmup-0.05-rouge-rougeL-temperature-1.0/critic/global_step_300/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..1c1d8d5c9024994f1d3b00f9662b8dd89ca13cf2 --- /dev/null +++ b/pilecc_cont1-r1-ppo-llama3.2-3b-it-em-warmup-0.05-rouge-rougeL-temperature-1.0/critic/global_step_300/tokenizer.json @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6b9e4e7fb171f92fd137b777cc2714bf87d11576700a1dcd7a399e7bbe39537b +size 17209920 diff --git a/pilecc_cont2-r1-ppo-llama3.2-3b-it-em-warmup-0.05-rouge-rougeL-temperature-1.0/actor/global_step_300/model-00001-of-00003.safetensors b/pilecc_cont2-r1-ppo-llama3.2-3b-it-em-warmup-0.05-rouge-rougeL-temperature-1.0/actor/global_step_300/model-00001-of-00003.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..8021d0ce6a8f91542fa941011a356dcbeb6637ff --- /dev/null +++ b/pilecc_cont2-r1-ppo-llama3.2-3b-it-em-warmup-0.05-rouge-rougeL-temperature-1.0/actor/global_step_300/model-00001-of-00003.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:463b39ebb21b685c68f9e3d7ea11d87d508a1919ecd6b5d7c5d63dc406ed98c9 +size 4998767360 diff --git a/pilecc_cont2-r1-ppo-llama3.2-3b-it-em-warmup-0.05-rouge-rougeL-temperature-1.0/actor/global_step_300/model-00002-of-00003.safetensors b/pilecc_cont2-r1-ppo-llama3.2-3b-it-em-warmup-0.05-rouge-rougeL-temperature-1.0/actor/global_step_300/model-00002-of-00003.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..de6221fa14b16a852e2a6a5e35ed7ddcc68457e2 --- /dev/null +++ b/pilecc_cont2-r1-ppo-llama3.2-3b-it-em-warmup-0.05-rouge-rougeL-temperature-1.0/actor/global_step_300/model-00002-of-00003.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ea5bcc96c231573edf69ae62f5388527982228a8824b9bdea23f9ad00886960e +size 4932808968 diff --git a/pilecc_cont2-r1-ppo-llama3.2-3b-it-em-warmup-0.05-rouge-rougeL-temperature-1.0/actor/global_step_300/model-00003-of-00003.safetensors b/pilecc_cont2-r1-ppo-llama3.2-3b-it-em-warmup-0.05-rouge-rougeL-temperature-1.0/actor/global_step_300/model-00003-of-00003.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..ec5ac3cb7bd4930a2591374dd9dfc0f4cadeb1e2 --- /dev/null +++ b/pilecc_cont2-r1-ppo-llama3.2-3b-it-em-warmup-0.05-rouge-rougeL-temperature-1.0/actor/global_step_300/model-00003-of-00003.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6d2938b4787359ce9e4a21e3d407813141ed1c7c50b7198c4db32a0ad1ac7e53 +size 4495462048 diff --git a/pilecc_cont2-r1-ppo-llama3.2-3b-it-em-warmup-0.05-rouge-rougeL-temperature-1.0/actor/global_step_300/tokenizer.json b/pilecc_cont2-r1-ppo-llama3.2-3b-it-em-warmup-0.05-rouge-rougeL-temperature-1.0/actor/global_step_300/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..1c1d8d5c9024994f1d3b00f9662b8dd89ca13cf2 --- /dev/null +++ b/pilecc_cont2-r1-ppo-llama3.2-3b-it-em-warmup-0.05-rouge-rougeL-temperature-1.0/actor/global_step_300/tokenizer.json @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6b9e4e7fb171f92fd137b777cc2714bf87d11576700a1dcd7a399e7bbe39537b +size 17209920 diff --git a/pilecc_cont2-r1-ppo-llama3.2-3b-it-em-warmup-0.05-rouge-rougeL-temperature-1.0/critic/global_step_300/model-00001-of-00003.safetensors b/pilecc_cont2-r1-ppo-llama3.2-3b-it-em-warmup-0.05-rouge-rougeL-temperature-1.0/critic/global_step_300/model-00001-of-00003.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..f2ac38192e3521280b46cd73a9ca4e02291756f8 --- /dev/null +++ b/pilecc_cont2-r1-ppo-llama3.2-3b-it-em-warmup-0.05-rouge-rougeL-temperature-1.0/critic/global_step_300/model-00001-of-00003.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6c2882c27152cc79b3f7949c172f6cda1a2b5b692ce13c03e3884afe0434afa8 +size 4998767360 diff --git a/pilecc_cont2-r1-ppo-llama3.2-3b-it-em-warmup-0.05-rouge-rougeL-temperature-1.0/critic/global_step_300/model-00002-of-00003.safetensors b/pilecc_cont2-r1-ppo-llama3.2-3b-it-em-warmup-0.05-rouge-rougeL-temperature-1.0/critic/global_step_300/model-00002-of-00003.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..6bb24cd995ddd55f086779844f5c7f2cbf7253b7 --- /dev/null +++ b/pilecc_cont2-r1-ppo-llama3.2-3b-it-em-warmup-0.05-rouge-rougeL-temperature-1.0/critic/global_step_300/model-00002-of-00003.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:606f1a9d5eb1df6a6e038aa04aa983a5e498f270008910f0639459623fc4fb15 +size 4932808968 diff --git a/pilecc_cont2-r1-ppo-llama3.2-3b-it-em-warmup-0.05-rouge-rougeL-temperature-1.0/critic/global_step_300/model-00003-of-00003.safetensors b/pilecc_cont2-r1-ppo-llama3.2-3b-it-em-warmup-0.05-rouge-rougeL-temperature-1.0/critic/global_step_300/model-00003-of-00003.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..df4c3068d1a3010480b8d7b738d361591c8b64c0 --- /dev/null +++ b/pilecc_cont2-r1-ppo-llama3.2-3b-it-em-warmup-0.05-rouge-rougeL-temperature-1.0/critic/global_step_300/model-00003-of-00003.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e20eaf8c3bb3c13fda4eb25152cbaf86cf745b6dfadddf69708e91cb48c6430b +size 2919464628 diff --git a/pilecc_cont2-r1-ppo-llama3.2-3b-it-em-warmup-0.05-rouge-rougeL-temperature-1.0/critic/global_step_300/tokenizer.json b/pilecc_cont2-r1-ppo-llama3.2-3b-it-em-warmup-0.05-rouge-rougeL-temperature-1.0/critic/global_step_300/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..1c1d8d5c9024994f1d3b00f9662b8dd89ca13cf2 --- /dev/null +++ b/pilecc_cont2-r1-ppo-llama3.2-3b-it-em-warmup-0.05-rouge-rougeL-temperature-1.0/critic/global_step_300/tokenizer.json @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6b9e4e7fb171f92fd137b777cc2714bf87d11576700a1dcd7a399e7bbe39537b +size 17209920