Upload checkpoints/math_operations/primitive_atomic_full_sft_then_2step_compositional_full_sft_t20260309

.gitattributes

@@ -72,3 +72,4 @@ checkpoints/math_operations/lora_sft_primitive_atomic_50k_t20260305/checkpoint-4
 checkpoints/math_operations/lora_sft_primitive_atomic_50k_t20260305/checkpoint-5150/tokenizer.json filter=lfs diff=lfs merge=lfs -text
 checkpoints/math_operations/primitive_atomic_full_sft_50k_lr1e4_t20260308/checkpoint-2082/tokenizer.json filter=lfs diff=lfs merge=lfs -text
 checkpoints/math_operations/primitive_atomic_full_sft_50k_lr1e4_t20260308/checkpoint-3123/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+checkpoints/math_operations/primitive_atomic_full_sft_then_2step_compositional_full_sft_t20260309/checkpoint-3125/tokenizer.json filter=lfs diff=lfs merge=lfs -text

checkpoints/math_operations/primitive_atomic_full_sft_then_2step_compositional_full_sft_t20260309/checkpoint-3125/chat_template.jinja

@@ -0,0 +1,4 @@
+{% if not add_generation_prompt is defined %}{% set add_generation_prompt = false %}{% endif %}{% for message in messages %}{{'<|im_start|>' + message['role'] + '
+' + message['content'] + '<|im_end|>' + '
+'}}{% endfor %}{% if add_generation_prompt %}{{ '<|im_start|>assistant
+' }}{% endif %}

checkpoints/math_operations/primitive_atomic_full_sft_then_2step_compositional_full_sft_t20260309/checkpoint-3125/config.json

@@ -0,0 +1,71 @@
+{
+  "architectures": [
+    "Qwen3ForCausalLM"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "bos_token_id": null,
+  "dtype": "bfloat16",
+  "eos_token_id": 151645,
+  "head_dim": 128,
+  "hidden_act": "silu",
+  "hidden_size": 2560,
+  "initializer_range": 0.02,
+  "intermediate_size": 9728,
+  "layer_types": [
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention"
+  ],
+  "max_position_embeddings": 262144,
+  "max_window_layers": 36,
+  "model_type": "qwen3",
+  "num_attention_heads": 32,
+  "num_hidden_layers": 36,
+  "num_key_value_heads": 8,
+  "pad_token_id": 151643,
+  "rms_norm_eps": 1e-06,
+  "rope_parameters": {
+    "rope_theta": 5000000,
+    "rope_type": "default"
+  },
+  "sliding_window": null,
+  "tie_word_embeddings": true,
+  "transformers_version": "5.0.0",
+  "use_cache": false,
+  "use_sliding_window": false,
+  "vocab_size": 151936
+}

checkpoints/math_operations/primitive_atomic_full_sft_then_2step_compositional_full_sft_t20260309/checkpoint-3125/generation_config.json

@@ -0,0 +1,12 @@
+{
+  "do_sample": true,
+  "eos_token_id": [
+    151645,
+    151643
+  ],
+  "pad_token_id": 151643,
+  "temperature": 0.7,
+  "top_k": 20,
+  "top_p": 0.8,
+  "transformers_version": "5.0.0"
+}

checkpoints/math_operations/primitive_atomic_full_sft_then_2step_compositional_full_sft_t20260309/checkpoint-3125/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:292d00bb09b3870f13e4d6c952cbf2bfa6917fe472738124a6ebb8df0796836c
+size 8822894520

checkpoints/math_operations/primitive_atomic_full_sft_then_2step_compositional_full_sft_t20260309/checkpoint-3125/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8780501c2f783665a77b85208cd3e759fe4f06e87bab00e2b7ca226cda4b40d1
+size 16090225449

checkpoints/math_operations/primitive_atomic_full_sft_then_2step_compositional_full_sft_t20260309/checkpoint-3125/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ed3753ab7977739b8eda494dd72defae5750f7283141b11a8f562160ba4c1a23
+size 14645

checkpoints/math_operations/primitive_atomic_full_sft_then_2step_compositional_full_sft_t20260309/checkpoint-3125/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dd5cdc84d2b4d9a932721799f114ebfcc55a4d590aebab27d5998a9f127ea471
+size 1465

checkpoints/math_operations/primitive_atomic_full_sft_then_2step_compositional_full_sft_t20260309/checkpoint-3125/tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:be75606093db2094d7cd20f3c2f385c212750648bd6ea4fb2bf507a6a4c55506
+size 11422650

checkpoints/math_operations/primitive_atomic_full_sft_then_2step_compositional_full_sft_t20260309/checkpoint-3125/tokenizer_config.json

@@ -0,0 +1,29 @@
+{
+  "add_prefix_space": false,
+  "backend": "tokenizers",
+  "bos_token": null,
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "errors": "replace",
+  "extra_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<|object_ref_start|>",
+    "<|object_ref_end|>",
+    "<|box_start|>",
+    "<|box_end|>",
+    "<|quad_start|>",
+    "<|quad_end|>",
+    "<|vision_start|>",
+    "<|vision_end|>",
+    "<|vision_pad|>",
+    "<|image_pad|>",
+    "<|video_pad|>"
+  ],
+  "is_local": true,
+  "model_max_length": 1010000,
+  "pad_token": "<|endoftext|>",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}

checkpoints/math_operations/primitive_atomic_full_sft_then_2step_compositional_full_sft_t20260309/checkpoint-3125/tokens_state.

@@ -0,0 +1 @@
+{"total": 51200000, "trainable": 5590610}

checkpoints/math_operations/primitive_atomic_full_sft_then_2step_compositional_full_sft_t20260309/checkpoint-3125/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0f19357fe7faa7293775d9e7faee6aabb3fe0214ae428e3c5d4fa8ace13d4ef3
+size 7185

checkpoints/math_operations/primitive_atomic_full_sft_then_2step_compositional_full_sft_t20260309/eval_results_easy_ops/eval_results.csv

@@ -0,0 +1,2 @@
+category,filename,total,correct,accuracy,format_found,format_accuracy,errors_count
+math_operations,balanced_test_alpaca_results,500,4,0.80,500,100.00,496

checkpoints/math_operations/primitive_atomic_full_sft_then_2step_compositional_full_sft_t20260309/eval_results_easy_ops/eval_summary.json

@@ -0,0 +1,133 @@
+{
+  "overall": {
+    "total": 500,
+    "correct": 4,
+    "accuracy": 0.8,
+    "format_found": 500,
+    "format_accuracy": 100.0
+  },
+  "per_operation": {
+    "a": {
+      "total": 25,
+      "correct": 0,
+      "accuracy": 0.0,
+      "format_found": 25
+    },
+    "b": {
+      "total": 25,
+      "correct": 0,
+      "accuracy": 0.0,
+      "format_found": 25
+    },
+    "c": {
+      "total": 25,
+      "correct": 1,
+      "accuracy": 4.0,
+      "format_found": 25
+    },
+    "d": {
+      "total": 25,
+      "correct": 0,
+      "accuracy": 0.0,
+      "format_found": 25
+    },
+    "e": {
+      "total": 25,
+      "correct": 0,
+      "accuracy": 0.0,
+      "format_found": 25
+    },
+    "f": {
+      "total": 25,
+      "correct": 0,
+      "accuracy": 0.0,
+      "format_found": 25
+    },
+    "g": {
+      "total": 25,
+      "correct": 0,
+      "accuracy": 0.0,
+      "format_found": 25
+    },
+    "h": {
+      "total": 25,
+      "correct": 0,
+      "accuracy": 0.0,
+      "format_found": 25
+    },
+    "i": {
+      "total": 25,
+      "correct": 0,
+      "accuracy": 0.0,
+      "format_found": 25
+    },
+    "j": {
+      "total": 25,
+      "correct": 0,
+      "accuracy": 0.0,
+      "format_found": 25
+    },
+    "k": {
+      "total": 25,
+      "correct": 0,
+      "accuracy": 0.0,
+      "format_found": 25
+    },
+    "l": {
+      "total": 25,
+      "correct": 0,
+      "accuracy": 0.0,
+      "format_found": 25
+    },
+    "m": {
+      "total": 25,
+      "correct": 0,
+      "accuracy": 0.0,
+      "format_found": 25
+    },
+    "n": {
+      "total": 25,
+      "correct": 1,
+      "accuracy": 4.0,
+      "format_found": 25
+    },
+    "o": {
+      "total": 25,
+      "correct": 0,
+      "accuracy": 0.0,
+      "format_found": 25
+    },
+    "p": {
+      "total": 25,
+      "correct": 0,
+      "accuracy": 0.0,
+      "format_found": 25
+    },
+    "q": {
+      "total": 25,
+      "correct": 0,
+      "accuracy": 0.0,
+      "format_found": 25
+    },
+    "r": {
+      "total": 25,
+      "correct": 0,
+      "accuracy": 0.0,
+      "format_found": 25
+    },
+    "s": {
+      "total": 25,
+      "correct": 0,
+      "accuracy": 0.0,
+      "format_found": 25
+    },
+    "t": {
+      "total": 25,
+      "correct": 2,
+      "accuracy": 8.0,
+      "format_found": 25
+    }
+  },
+  "n_errors": 496,
+  "results_file": "/home/jiaruil5/math_rl/mix_teachers/r3lit_rl/mix_teachers/checkpoints/math_operations/primitive_atomic_full_sft_then_2step_compositional_full_sft_t20260309/eval_results_easy_ops/balanced_test_alpaca_results.jsonl"
+}