instance_id stringlengths 12 12 | source_index int64 0 40.1k | num_jobs int64 2 7 | num_machines int64 2 6 | total_steps int64 4 36 | step_idx int64 0 35 | state_json dict | state_text stringlengths 1.22k 3.37k | problem_context_text stringlengths 176 697 | feasible_jobs listlengths 1 7 | feasible_action_codes listlengths 1 7 | action_code_to_job unknown | teacher_operation_idx int64 0 5 | teacher_machine int64 0 5 | env_mode stringclasses 1
value | dispatch_teacher_exact bool 2
classes | feature_schema_version stringclasses 1
value | selected_job int64 0 6 | selected_action_code stringlengths 7 7 | reason_input_text stringlengths 2.05k 4.83k | reason_target_text stringlengths 411 2.05k | chosen_transition_features dict | contrast_action_codes listlengths 0 3 | contrast_transition_features listlengths 0 3 | reason_source stringclasses 1
value |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
train_000000 | 0 | 2 | 2 | 4 | 0 | {
"step_idx": 0,
"total_steps": 4,
"scheduled_ratio": 0,
"current_time": 0,
"current_cmax": 0,
"job_next_op": [
0,
0
],
"job_total_ops": [
2,
2
],
"job_total_work": [
49,
53
],
"job_ready_time": [
0,
0
],
"machine_ready_time": [
0,
0
],
"next_machi... | You are solving JSSP with event-driven dispatching.
Objective: minimize final makespan (Cmax) while respecting precedence and machine availability.
Static problem context:
Problem: 2 jobs x 2 machines (total_ops=4)
Fixed operation routes:
Job 0: ops=2, total_work=49, route=(M0,t21) -> (M1,t28)
Job 1: ops=2, total_work=... | Problem: 2 jobs x 2 machines (total_ops=4)
Fixed operation routes:
Job 0: ops=2, total_work=49, route=(M0,t21) -> (M1,t28)
Job 1: ops=2, total_work=53, route=(M1,t26) -> (M0,t27) | [
0,
1
] | [
"<A0410>",
"<A4507>"
] | {
"<A0410>": 1,
"<A4507>": 0
} | 0 | 0 | dispatch | true | jssp_step_reason_dispatch_v1_action_token | 0 | <A4507> | You are analyzing an already-selected JSSP action.
Objective: explain why this action was selected and why strong alternatives were not selected.
Ground the explanation in explicit evidence only: immediate timing/Cmax impact, bottleneck-machine use or release, downstream route exposure (next2), waiting/idle trade-offs,... | Reason: <A4507> is feasible on M0 (est_start=0, est_end=21). Projected Cmax changes 0->21 (delta=21). It reduces this job from remaining_work 49->28 and remaining_ops 2->1, while exposing the next route on M1 (t=28). It also exposes a downstream step on bottleneck M1 (next2_p=28), pulling future critical work closer. I... | {
"action_code": "<A4507>",
"job_id": 0,
"machine_id": 0,
"machine_token": "M0",
"proc_time": 21,
"next_machine": 0,
"next_proc_time": 21,
"next2_machine": 1,
"next2_proc_time": 28,
"remaining_ops_before": 2,
"remaining_ops_after": 1,
"remaining_work_before": 49,
"remaining_work_after": 28,
... | [
"<A0410>"
] | [
{
"action_code": "<A0410>",
"job_id": 1,
"machine_id": 1,
"machine_token": "M1",
"proc_time": 26,
"next_machine": 1,
"next_proc_time": 26,
"next2_machine": 0,
"next2_proc_time": 27,
"remaining_ops_before": 2,
"remaining_ops_after": 1,
"remaining_work_before": 53,
... | deterministic_teacher_v1 |
train_000000 | 0 | 2 | 2 | 4 | 1 | {
"step_idx": 1,
"total_steps": 4,
"scheduled_ratio": 0.25,
"current_time": 0,
"current_cmax": 21,
"job_next_op": [
1,
0
],
"job_total_ops": [
2,
2
],
"job_total_work": [
49,
53
],
"job_ready_time": [
21,
0
],
"machine_ready_time": [
21,
0
],
"next... | You are solving JSSP with event-driven dispatching.
Objective: minimize final makespan (Cmax) while respecting precedence and machine availability.
Static problem context:
Problem: 2 jobs x 2 machines (total_ops=4)
Fixed operation routes:
Job 0: ops=2, total_work=49, route=(M0,t21) -> (M1,t28)
Job 1: ops=2, total_work=... | Problem: 2 jobs x 2 machines (total_ops=4)
Fixed operation routes:
Job 0: ops=2, total_work=49, route=(M0,t21) -> (M1,t28)
Job 1: ops=2, total_work=53, route=(M1,t26) -> (M0,t27) | [
1
] | [
"<A0632>"
] | {
"<A0632>": 1
} | 0 | 1 | dispatch | true | jssp_step_reason_dispatch_v1_action_token | 1 | <A0632> | You are analyzing an already-selected JSSP action.
Objective: explain why this action was selected and why strong alternatives were not selected.
Ground the explanation in explicit evidence only: immediate timing/Cmax impact, bottleneck-machine use or release, downstream route exposure (next2), waiting/idle trade-offs,... | Reason: <A0632> is feasible on M1 (est_start=0, est_end=26). Projected Cmax changes 21->26 (delta=5). It reduces this job from remaining_work 53->27 and remaining_ops 2->1, while exposing the next route on M0 (t=27). It directly activates the current bottleneck machine M1 (remaining_load=54, ops_left=2), so delaying th... | {
"action_code": "<A0632>",
"job_id": 1,
"machine_id": 1,
"machine_token": "M1",
"proc_time": 26,
"next_machine": 1,
"next_proc_time": 26,
"next2_machine": 0,
"next2_proc_time": 27,
"remaining_ops_before": 2,
"remaining_ops_after": 1,
"remaining_work_before": 53,
"remaining_work_after": 27,
... | [] | [] | deterministic_teacher_v1 |
train_000000 | 0 | 2 | 2 | 4 | 2 | {
"step_idx": 2,
"total_steps": 4,
"scheduled_ratio": 0.5,
"current_time": 26,
"current_cmax": 26,
"job_next_op": [
1,
1
],
"job_total_ops": [
2,
2
],
"job_total_work": [
49,
53
],
"job_ready_time": [
21,
26
],
"machine_ready_time": [
21,
26
],
"ne... | You are solving JSSP with event-driven dispatching.
Objective: minimize final makespan (Cmax) while respecting precedence and machine availability.
Static problem context:
Problem: 2 jobs x 2 machines (total_ops=4)
Fixed operation routes:
Job 0: ops=2, total_work=49, route=(M0,t21) -> (M1,t28)
Job 1: ops=2, total_work=... | Problem: 2 jobs x 2 machines (total_ops=4)
Fixed operation routes:
Job 0: ops=2, total_work=49, route=(M0,t21) -> (M1,t28)
Job 1: ops=2, total_work=53, route=(M1,t26) -> (M0,t27) | [
0,
1
] | [
"<A8521>",
"<A8880>"
] | {
"<A8521>": 0,
"<A8880>": 1
} | 1 | 0 | dispatch | true | jssp_step_reason_dispatch_v1_action_token | 1 | <A8880> | You are analyzing an already-selected JSSP action.
Objective: explain why this action was selected and why strong alternatives were not selected.
Ground the explanation in explicit evidence only: immediate timing/Cmax impact, bottleneck-machine use or release, downstream route exposure (next2), waiting/idle trade-offs,... | Reason: <A8880> is feasible on M0 (est_start=26, est_end=53). Projected Cmax changes 26->53 (delta=27). It completes this job, eliminating the remaining work from 27 to 0. It works on high-load machine M0 (remaining_load=27), which is close to the current bottleneck pressure. It fills machine idle gap=5 on M0, which he... | {
"action_code": "<A8880>",
"job_id": 1,
"machine_id": 0,
"machine_token": "M0",
"proc_time": 27,
"next_machine": 0,
"next_proc_time": 27,
"next2_machine": -1,
"next2_proc_time": 0,
"remaining_ops_before": 1,
"remaining_ops_after": 0,
"remaining_work_before": 27,
"remaining_work_after": 0,
"... | [
"<A8521>"
] | [
{
"action_code": "<A8521>",
"job_id": 0,
"machine_id": 1,
"machine_token": "M1",
"proc_time": 28,
"next_machine": 1,
"next_proc_time": 28,
"next2_machine": -1,
"next2_proc_time": 0,
"remaining_ops_before": 1,
"remaining_ops_after": 0,
"remaining_work_before": 28,
... | deterministic_teacher_v1 |
train_000000 | 0 | 2 | 2 | 4 | 3 | {
"step_idx": 3,
"total_steps": 4,
"scheduled_ratio": 0.75,
"current_time": 26,
"current_cmax": 53,
"job_next_op": [
1,
2
],
"job_total_ops": [
2,
2
],
"job_total_work": [
49,
53
],
"job_ready_time": [
21,
53
],
"machine_ready_time": [
53,
26
],
"n... | You are solving JSSP with event-driven dispatching.
Objective: minimize final makespan (Cmax) while respecting precedence and machine availability.
Static problem context:
Problem: 2 jobs x 2 machines (total_ops=4)
Fixed operation routes:
Job 0: ops=2, total_work=49, route=(M0,t21) -> (M1,t28)
Job 1: ops=2, total_work=... | Problem: 2 jobs x 2 machines (total_ops=4)
Fixed operation routes:
Job 0: ops=2, total_work=49, route=(M0,t21) -> (M1,t28)
Job 1: ops=2, total_work=53, route=(M1,t26) -> (M0,t27) | [
0
] | [
"<A4455>"
] | {
"<A4455>": 0
} | 1 | 1 | dispatch | true | jssp_step_reason_dispatch_v1_action_token | 0 | <A4455> | You are analyzing an already-selected JSSP action.
Objective: explain why this action was selected and why strong alternatives were not selected.
Ground the explanation in explicit evidence only: immediate timing/Cmax impact, bottleneck-machine use or release, downstream route exposure (next2), waiting/idle trade-offs,... | Reason: <A4455> is feasible on M1 (est_start=26, est_end=54). Projected Cmax changes 53->54 (delta=1). It completes this job, eliminating the remaining work from 28 to 0. It directly activates the current bottleneck machine M1 (remaining_load=28, ops_left=1), so delaying this move would postpone work on the heaviest un... | {
"action_code": "<A4455>",
"job_id": 0,
"machine_id": 1,
"machine_token": "M1",
"proc_time": 28,
"next_machine": 1,
"next_proc_time": 28,
"next2_machine": -1,
"next2_proc_time": 0,
"remaining_ops_before": 1,
"remaining_ops_after": 0,
"remaining_work_before": 28,
"remaining_work_after": 0,
"... | [] | [] | deterministic_teacher_v1 |
train_000001 | 1 | 2 | 2 | 4 | 0 | {
"step_idx": 0,
"total_steps": 4,
"scheduled_ratio": 0,
"current_time": 0,
"current_cmax": 0,
"job_next_op": [
0,
0
],
"job_total_ops": [
2,
2
],
"job_total_work": [
40,
42
],
"job_ready_time": [
0,
0
],
"machine_ready_time": [
0,
0
],
"next_machi... | You are solving JSSP with event-driven dispatching.
Objective: minimize final makespan (Cmax) while respecting precedence and machine availability.
Static problem context:
Problem: 2 jobs x 2 machines (total_ops=4)
Fixed operation routes:
Job 0: ops=2, total_work=40, route=(M0,t21) -> (M1,t19)
Job 1: ops=2, total_work=... | Problem: 2 jobs x 2 machines (total_ops=4)
Fixed operation routes:
Job 0: ops=2, total_work=40, route=(M0,t21) -> (M1,t19)
Job 1: ops=2, total_work=42, route=(M0,t28) -> (M1,t14) | [
0,
1
] | [
"<A6685>",
"<A8174>"
] | {
"<A6685>": 0,
"<A8174>": 1
} | 0 | 0 | dispatch | true | jssp_step_reason_dispatch_v1_action_token | 0 | <A6685> | You are analyzing an already-selected JSSP action.
Objective: explain why this action was selected and why strong alternatives were not selected.
Ground the explanation in explicit evidence only: immediate timing/Cmax impact, bottleneck-machine use or release, downstream route exposure (next2), waiting/idle trade-offs,... | Reason: <A6685> is feasible on M0 (est_start=0, est_end=21). Projected Cmax changes 0->21 (delta=21). It reduces this job from remaining_work 40->19 and remaining_ops 2->1, while exposing the next route on M1 (t=19). It directly activates the current bottleneck machine M0 (remaining_load=49, ops_left=2), so delaying th... | {
"action_code": "<A6685>",
"job_id": 0,
"machine_id": 0,
"machine_token": "M0",
"proc_time": 21,
"next_machine": 0,
"next_proc_time": 21,
"next2_machine": 1,
"next2_proc_time": 19,
"remaining_ops_before": 2,
"remaining_ops_after": 1,
"remaining_work_before": 40,
"remaining_work_after": 19,
... | [
"<A8174>"
] | [
{
"action_code": "<A8174>",
"job_id": 1,
"machine_id": 0,
"machine_token": "M0",
"proc_time": 28,
"next_machine": 0,
"next_proc_time": 28,
"next2_machine": 1,
"next2_proc_time": 14,
"remaining_ops_before": 2,
"remaining_ops_after": 1,
"remaining_work_before": 42,
... | deterministic_teacher_v1 |
train_000001 | 1 | 2 | 2 | 4 | 1 | {
"step_idx": 1,
"total_steps": 4,
"scheduled_ratio": 0.25,
"current_time": 21,
"current_cmax": 21,
"job_next_op": [
1,
0
],
"job_total_ops": [
2,
2
],
"job_total_work": [
40,
42
],
"job_ready_time": [
21,
0
],
"machine_ready_time": [
21,
0
],
"nex... | You are solving JSSP with event-driven dispatching.
Objective: minimize final makespan (Cmax) while respecting precedence and machine availability.
Static problem context:
Problem: 2 jobs x 2 machines (total_ops=4)
Fixed operation routes:
Job 0: ops=2, total_work=40, route=(M0,t21) -> (M1,t19)
Job 1: ops=2, total_work=... | Problem: 2 jobs x 2 machines (total_ops=4)
Fixed operation routes:
Job 0: ops=2, total_work=40, route=(M0,t21) -> (M1,t19)
Job 1: ops=2, total_work=42, route=(M0,t28) -> (M1,t14) | [
0,
1
] | [
"<A5740>",
"<A1177>"
] | {
"<A5740>": 0,
"<A1177>": 1
} | 0 | 0 | dispatch | true | jssp_step_reason_dispatch_v1_action_token | 1 | <A1177> | You are analyzing an already-selected JSSP action.
Objective: explain why this action was selected and why strong alternatives were not selected.
Ground the explanation in explicit evidence only: immediate timing/Cmax impact, bottleneck-machine use or release, downstream route exposure (next2), waiting/idle trade-offs,... | Reason: <A1177> is feasible on M0 (est_start=21, est_end=49). Projected Cmax changes 21->49 (delta=28). It reduces this job from remaining_work 42->14 and remaining_ops 2->1, while exposing the next route on M1 (t=14). It also exposes a downstream step on bottleneck M1 (next2_p=14), pulling future critical work closer.... | {
"action_code": "<A1177>",
"job_id": 1,
"machine_id": 0,
"machine_token": "M0",
"proc_time": 28,
"next_machine": 0,
"next_proc_time": 28,
"next2_machine": 1,
"next2_proc_time": 14,
"remaining_ops_before": 2,
"remaining_ops_after": 1,
"remaining_work_before": 42,
"remaining_work_after": 14,
... | [
"<A5740>"
] | [
{
"action_code": "<A5740>",
"job_id": 0,
"machine_id": 1,
"machine_token": "M1",
"proc_time": 19,
"next_machine": 1,
"next_proc_time": 19,
"next2_machine": -1,
"next2_proc_time": 0,
"remaining_ops_before": 1,
"remaining_ops_after": 0,
"remaining_work_before": 19,
... | deterministic_teacher_v1 |
train_000001 | 1 | 2 | 2 | 4 | 2 | {
"step_idx": 2,
"total_steps": 4,
"scheduled_ratio": 0.5,
"current_time": 21,
"current_cmax": 49,
"job_next_op": [
1,
1
],
"job_total_ops": [
2,
2
],
"job_total_work": [
40,
42
],
"job_ready_time": [
21,
49
],
"machine_ready_time": [
49,
0
],
"nex... | You are solving JSSP with event-driven dispatching.
Objective: minimize final makespan (Cmax) while respecting precedence and machine availability.
Static problem context:
Problem: 2 jobs x 2 machines (total_ops=4)
Fixed operation routes:
Job 0: ops=2, total_work=40, route=(M0,t21) -> (M1,t19)
Job 1: ops=2, total_work=... | Problem: 2 jobs x 2 machines (total_ops=4)
Fixed operation routes:
Job 0: ops=2, total_work=40, route=(M0,t21) -> (M1,t19)
Job 1: ops=2, total_work=42, route=(M0,t28) -> (M1,t14) | [
0
] | [
"<A8641>"
] | {
"<A8641>": 0
} | 1 | 1 | dispatch | true | jssp_step_reason_dispatch_v1_action_token | 0 | <A8641> | You are analyzing an already-selected JSSP action.
Objective: explain why this action was selected and why strong alternatives were not selected.
Ground the explanation in explicit evidence only: immediate timing/Cmax impact, bottleneck-machine use or release, downstream route exposure (next2), waiting/idle trade-offs,... | Reason: <A8641> is feasible on M1 (est_start=21, est_end=40). Projected Cmax changes 49->49 (delta=0). It completes this job, eliminating the remaining work from 19 to 0. It directly activates the current bottleneck machine M1 (remaining_load=33, ops_left=2), so delaying this move would postpone work on the heaviest un... | {
"action_code": "<A8641>",
"job_id": 0,
"machine_id": 1,
"machine_token": "M1",
"proc_time": 19,
"next_machine": 1,
"next_proc_time": 19,
"next2_machine": -1,
"next2_proc_time": 0,
"remaining_ops_before": 1,
"remaining_ops_after": 0,
"remaining_work_before": 19,
"remaining_work_after": 0,
"... | [] | [] | deterministic_teacher_v1 |
train_000001 | 1 | 2 | 2 | 4 | 3 | {
"step_idx": 3,
"total_steps": 4,
"scheduled_ratio": 0.75,
"current_time": 49,
"current_cmax": 49,
"job_next_op": [
2,
1
],
"job_total_ops": [
2,
2
],
"job_total_work": [
40,
42
],
"job_ready_time": [
40,
49
],
"machine_ready_time": [
49,
40
],
"n... | You are solving JSSP with event-driven dispatching.
Objective: minimize final makespan (Cmax) while respecting precedence and machine availability.
Static problem context:
Problem: 2 jobs x 2 machines (total_ops=4)
Fixed operation routes:
Job 0: ops=2, total_work=40, route=(M0,t21) -> (M1,t19)
Job 1: ops=2, total_work=... | Problem: 2 jobs x 2 machines (total_ops=4)
Fixed operation routes:
Job 0: ops=2, total_work=40, route=(M0,t21) -> (M1,t19)
Job 1: ops=2, total_work=42, route=(M0,t28) -> (M1,t14) | [
1
] | [
"<A4673>"
] | {
"<A4673>": 1
} | 1 | 1 | dispatch | true | jssp_step_reason_dispatch_v1_action_token | 1 | <A4673> | You are analyzing an already-selected JSSP action.
Objective: explain why this action was selected and why strong alternatives were not selected.
Ground the explanation in explicit evidence only: immediate timing/Cmax impact, bottleneck-machine use or release, downstream route exposure (next2), waiting/idle trade-offs,... | Reason: <A4673> is feasible on M1 (est_start=49, est_end=63). Projected Cmax changes 49->63 (delta=14). It completes this job, eliminating the remaining work from 14 to 0. It directly activates the current bottleneck machine M1 (remaining_load=14, ops_left=1), so delaying this move would postpone work on the heaviest u... | {
"action_code": "<A4673>",
"job_id": 1,
"machine_id": 1,
"machine_token": "M1",
"proc_time": 14,
"next_machine": 1,
"next_proc_time": 14,
"next2_machine": -1,
"next2_proc_time": 0,
"remaining_ops_before": 1,
"remaining_ops_after": 0,
"remaining_work_before": 14,
"remaining_work_after": 0,
"... | [] | [] | deterministic_teacher_v1 |
train_000002 | 2 | 2 | 2 | 4 | 0 | {
"step_idx": 0,
"total_steps": 4,
"scheduled_ratio": 0,
"current_time": 0,
"current_cmax": 0,
"job_next_op": [
0,
0
],
"job_total_ops": [
2,
2
],
"job_total_work": [
51,
32
],
"job_ready_time": [
0,
0
],
"machine_ready_time": [
0,
0
],
"next_machi... | You are solving JSSP with event-driven dispatching.
Objective: minimize final makespan (Cmax) while respecting precedence and machine availability.
Static problem context:
Problem: 2 jobs x 2 machines (total_ops=4)
Fixed operation routes:
Job 0: ops=2, total_work=51, route=(M1,t23) -> (M0,t28)
Job 1: ops=2, total_work=... | Problem: 2 jobs x 2 machines (total_ops=4)
Fixed operation routes:
Job 0: ops=2, total_work=51, route=(M1,t23) -> (M0,t28)
Job 1: ops=2, total_work=32, route=(M0,t17) -> (M1,t15) | [
0,
1
] | [
"<A6375>",
"<A1051>"
] | {
"<A6375>": 1,
"<A1051>": 0
} | 0 | 0 | dispatch | true | jssp_step_reason_dispatch_v1_action_token | 1 | <A6375> | You are analyzing an already-selected JSSP action.
Objective: explain why this action was selected and why strong alternatives were not selected.
Ground the explanation in explicit evidence only: immediate timing/Cmax impact, bottleneck-machine use or release, downstream route exposure (next2), waiting/idle trade-offs,... | Reason: <A6375> is feasible on M0 (est_start=0, est_end=17). Projected Cmax changes 0->17 (delta=17). It reduces this job from remaining_work 32->15 and remaining_ops 2->1, while exposing the next route on M1 (t=15). It directly activates the current bottleneck machine M0 (remaining_load=45, ops_left=2), so delaying th... | {
"action_code": "<A6375>",
"job_id": 1,
"machine_id": 0,
"machine_token": "M0",
"proc_time": 17,
"next_machine": 0,
"next_proc_time": 17,
"next2_machine": 1,
"next2_proc_time": 15,
"remaining_ops_before": 2,
"remaining_ops_after": 1,
"remaining_work_before": 32,
"remaining_work_after": 15,
... | [
"<A1051>"
] | [
{
"action_code": "<A1051>",
"job_id": 0,
"machine_id": 1,
"machine_token": "M1",
"proc_time": 23,
"next_machine": 1,
"next_proc_time": 23,
"next2_machine": 0,
"next2_proc_time": 28,
"remaining_ops_before": 2,
"remaining_ops_after": 1,
"remaining_work_before": 51,
... | deterministic_teacher_v1 |
train_000002 | 2 | 2 | 2 | 4 | 1 | {
"step_idx": 1,
"total_steps": 4,
"scheduled_ratio": 0.25,
"current_time": 0,
"current_cmax": 17,
"job_next_op": [
0,
1
],
"job_total_ops": [
2,
2
],
"job_total_work": [
51,
32
],
"job_ready_time": [
0,
17
],
"machine_ready_time": [
17,
0
],
"next... | You are solving JSSP with event-driven dispatching.
Objective: minimize final makespan (Cmax) while respecting precedence and machine availability.
Static problem context:
Problem: 2 jobs x 2 machines (total_ops=4)
Fixed operation routes:
Job 0: ops=2, total_work=51, route=(M1,t23) -> (M0,t28)
Job 1: ops=2, total_work=... | Problem: 2 jobs x 2 machines (total_ops=4)
Fixed operation routes:
Job 0: ops=2, total_work=51, route=(M1,t23) -> (M0,t28)
Job 1: ops=2, total_work=32, route=(M0,t17) -> (M1,t15) | [
0
] | [
"<A1048>"
] | {
"<A1048>": 0
} | 0 | 1 | dispatch | true | jssp_step_reason_dispatch_v1_action_token | 0 | <A1048> | You are analyzing an already-selected JSSP action.
Objective: explain why this action was selected and why strong alternatives were not selected.
Ground the explanation in explicit evidence only: immediate timing/Cmax impact, bottleneck-machine use or release, downstream route exposure (next2), waiting/idle trade-offs,... | Reason: <A1048> is feasible on M1 (est_start=0, est_end=23). Projected Cmax changes 17->23 (delta=6). It reduces this job from remaining_work 51->28 and remaining_ops 2->1, while exposing the next route on M0 (t=28). It directly activates the current bottleneck machine M1 (remaining_load=38, ops_left=2), so delaying th... | {
"action_code": "<A1048>",
"job_id": 0,
"machine_id": 1,
"machine_token": "M1",
"proc_time": 23,
"next_machine": 1,
"next_proc_time": 23,
"next2_machine": 0,
"next2_proc_time": 28,
"remaining_ops_before": 2,
"remaining_ops_after": 1,
"remaining_work_before": 51,
"remaining_work_after": 28,
... | [] | [] | deterministic_teacher_v1 |
train_000002 | 2 | 2 | 2 | 4 | 2 | {
"step_idx": 2,
"total_steps": 4,
"scheduled_ratio": 0.5,
"current_time": 23,
"current_cmax": 23,
"job_next_op": [
1,
1
],
"job_total_ops": [
2,
2
],
"job_total_work": [
51,
32
],
"job_ready_time": [
23,
17
],
"machine_ready_time": [
17,
23
],
"ne... | You are solving JSSP with event-driven dispatching.
Objective: minimize final makespan (Cmax) while respecting precedence and machine availability.
Static problem context:
Problem: 2 jobs x 2 machines (total_ops=4)
Fixed operation routes:
Job 0: ops=2, total_work=51, route=(M1,t23) -> (M0,t28)
Job 1: ops=2, total_work=... | Problem: 2 jobs x 2 machines (total_ops=4)
Fixed operation routes:
Job 0: ops=2, total_work=51, route=(M1,t23) -> (M0,t28)
Job 1: ops=2, total_work=32, route=(M0,t17) -> (M1,t15) | [
0,
1
] | [
"<A3040>",
"<A2900>"
] | {
"<A3040>": 0,
"<A2900>": 1
} | 1 | 0 | dispatch | true | jssp_step_reason_dispatch_v1_action_token | 0 | <A3040> | You are analyzing an already-selected JSSP action.
Objective: explain why this action was selected and why strong alternatives were not selected.
Ground the explanation in explicit evidence only: immediate timing/Cmax impact, bottleneck-machine use or release, downstream route exposure (next2), waiting/idle trade-offs,... | Reason: <A3040> is feasible on M0 (est_start=23, est_end=51). Projected Cmax changes 23->51 (delta=28). It completes this job, eliminating the remaining work from 28 to 0. It directly activates the current bottleneck machine M0 (remaining_load=28, ops_left=1), so delaying this move would postpone work on the heaviest u... | {
"action_code": "<A3040>",
"job_id": 0,
"machine_id": 0,
"machine_token": "M0",
"proc_time": 28,
"next_machine": 0,
"next_proc_time": 28,
"next2_machine": -1,
"next2_proc_time": 0,
"remaining_ops_before": 1,
"remaining_ops_after": 0,
"remaining_work_before": 28,
"remaining_work_after": 0,
"... | [
"<A2900>"
] | [
{
"action_code": "<A2900>",
"job_id": 1,
"machine_id": 1,
"machine_token": "M1",
"proc_time": 15,
"next_machine": 1,
"next_proc_time": 15,
"next2_machine": -1,
"next2_proc_time": 0,
"remaining_ops_before": 1,
"remaining_ops_after": 0,
"remaining_work_before": 15,
... | deterministic_teacher_v1 |
train_000002 | 2 | 2 | 2 | 4 | 3 | {
"step_idx": 3,
"total_steps": 4,
"scheduled_ratio": 0.75,
"current_time": 23,
"current_cmax": 51,
"job_next_op": [
2,
1
],
"job_total_ops": [
2,
2
],
"job_total_work": [
51,
32
],
"job_ready_time": [
51,
17
],
"machine_ready_time": [
51,
23
],
"n... | You are solving JSSP with event-driven dispatching.
Objective: minimize final makespan (Cmax) while respecting precedence and machine availability.
Static problem context:
Problem: 2 jobs x 2 machines (total_ops=4)
Fixed operation routes:
Job 0: ops=2, total_work=51, route=(M1,t23) -> (M0,t28)
Job 1: ops=2, total_work=... | Problem: 2 jobs x 2 machines (total_ops=4)
Fixed operation routes:
Job 0: ops=2, total_work=51, route=(M1,t23) -> (M0,t28)
Job 1: ops=2, total_work=32, route=(M0,t17) -> (M1,t15) | [
1
] | [
"<A5342>"
] | {
"<A5342>": 1
} | 1 | 1 | dispatch | true | jssp_step_reason_dispatch_v1_action_token | 1 | <A5342> | You are analyzing an already-selected JSSP action.
Objective: explain why this action was selected and why strong alternatives were not selected.
Ground the explanation in explicit evidence only: immediate timing/Cmax impact, bottleneck-machine use or release, downstream route exposure (next2), waiting/idle trade-offs,... | Reason: <A5342> is feasible on M1 (est_start=23, est_end=38). Projected Cmax changes 51->51 (delta=0). It completes this job, eliminating the remaining work from 15 to 0. It directly activates the current bottleneck machine M1 (remaining_load=15, ops_left=1), so delaying this move would postpone work on the heaviest un... | {
"action_code": "<A5342>",
"job_id": 1,
"machine_id": 1,
"machine_token": "M1",
"proc_time": 15,
"next_machine": 1,
"next_proc_time": 15,
"next2_machine": -1,
"next2_proc_time": 0,
"remaining_ops_before": 1,
"remaining_ops_after": 0,
"remaining_work_before": 15,
"remaining_work_after": 0,
"... | [] | [] | deterministic_teacher_v1 |
train_000003 | 3 | 2 | 2 | 4 | 0 | {
"step_idx": 0,
"total_steps": 4,
"scheduled_ratio": 0,
"current_time": 0,
"current_cmax": 0,
"job_next_op": [
0,
0
],
"job_total_ops": [
2,
2
],
"job_total_work": [
49,
40
],
"job_ready_time": [
0,
0
],
"machine_ready_time": [
0,
0
],
"next_machi... | You are solving JSSP with event-driven dispatching.
Objective: minimize final makespan (Cmax) while respecting precedence and machine availability.
Static problem context:
Problem: 2 jobs x 2 machines (total_ops=4)
Fixed operation routes:
Job 0: ops=2, total_work=49, route=(M0,t30) -> (M1,t19)
Job 1: ops=2, total_work=... | Problem: 2 jobs x 2 machines (total_ops=4)
Fixed operation routes:
Job 0: ops=2, total_work=49, route=(M0,t30) -> (M1,t19)
Job 1: ops=2, total_work=40, route=(M1,t22) -> (M0,t18) | [
0,
1
] | [
"<A7732>",
"<A9303>"
] | {
"<A7732>": 1,
"<A9303>": 0
} | 0 | 0 | dispatch | true | jssp_step_reason_dispatch_v1_action_token | 0 | <A9303> | You are analyzing an already-selected JSSP action.
Objective: explain why this action was selected and why strong alternatives were not selected.
Ground the explanation in explicit evidence only: immediate timing/Cmax impact, bottleneck-machine use or release, downstream route exposure (next2), waiting/idle trade-offs,... | Reason: <A9303> is feasible on M0 (est_start=0, est_end=30). Projected Cmax changes 0->30 (delta=30). It reduces this job from remaining_work 49->19 and remaining_ops 2->1, while exposing the next route on M1 (t=19). It directly activates the current bottleneck machine M0 (remaining_load=48, ops_left=2), so delaying th... | {
"action_code": "<A9303>",
"job_id": 0,
"machine_id": 0,
"machine_token": "M0",
"proc_time": 30,
"next_machine": 0,
"next_proc_time": 30,
"next2_machine": 1,
"next2_proc_time": 19,
"remaining_ops_before": 2,
"remaining_ops_after": 1,
"remaining_work_before": 49,
"remaining_work_after": 19,
... | [
"<A7732>"
] | [
{
"action_code": "<A7732>",
"job_id": 1,
"machine_id": 1,
"machine_token": "M1",
"proc_time": 22,
"next_machine": 1,
"next_proc_time": 22,
"next2_machine": 0,
"next2_proc_time": 18,
"remaining_ops_before": 2,
"remaining_ops_after": 1,
"remaining_work_before": 40,
... | deterministic_teacher_v1 |
train_000003 | 3 | 2 | 2 | 4 | 1 | {
"step_idx": 1,
"total_steps": 4,
"scheduled_ratio": 0.25,
"current_time": 0,
"current_cmax": 30,
"job_next_op": [
1,
0
],
"job_total_ops": [
2,
2
],
"job_total_work": [
49,
40
],
"job_ready_time": [
30,
0
],
"machine_ready_time": [
30,
0
],
"next... | You are solving JSSP with event-driven dispatching.
Objective: minimize final makespan (Cmax) while respecting precedence and machine availability.
Static problem context:
Problem: 2 jobs x 2 machines (total_ops=4)
Fixed operation routes:
Job 0: ops=2, total_work=49, route=(M0,t30) -> (M1,t19)
Job 1: ops=2, total_work=... | Problem: 2 jobs x 2 machines (total_ops=4)
Fixed operation routes:
Job 0: ops=2, total_work=49, route=(M0,t30) -> (M1,t19)
Job 1: ops=2, total_work=40, route=(M1,t22) -> (M0,t18) | [
1
] | [
"<A1117>"
] | {
"<A1117>": 1
} | 0 | 1 | dispatch | true | jssp_step_reason_dispatch_v1_action_token | 1 | <A1117> | You are analyzing an already-selected JSSP action.
Objective: explain why this action was selected and why strong alternatives were not selected.
Ground the explanation in explicit evidence only: immediate timing/Cmax impact, bottleneck-machine use or release, downstream route exposure (next2), waiting/idle trade-offs,... | Reason: <A1117> is feasible on M1 (est_start=0, est_end=22). Projected Cmax changes 30->30 (delta=0). It reduces this job from remaining_work 40->18 and remaining_ops 2->1, while exposing the next route on M0 (t=18). It directly activates the current bottleneck machine M1 (remaining_load=41, ops_left=2), so delaying th... | {
"action_code": "<A1117>",
"job_id": 1,
"machine_id": 1,
"machine_token": "M1",
"proc_time": 22,
"next_machine": 1,
"next_proc_time": 22,
"next2_machine": 0,
"next2_proc_time": 18,
"remaining_ops_before": 2,
"remaining_ops_after": 1,
"remaining_work_before": 40,
"remaining_work_after": 18,
... | [] | [] | deterministic_teacher_v1 |
train_000003 | 3 | 2 | 2 | 4 | 2 | {
"step_idx": 2,
"total_steps": 4,
"scheduled_ratio": 0.5,
"current_time": 30,
"current_cmax": 30,
"job_next_op": [
1,
1
],
"job_total_ops": [
2,
2
],
"job_total_work": [
49,
40
],
"job_ready_time": [
30,
22
],
"machine_ready_time": [
30,
22
],
"ne... | You are solving JSSP with event-driven dispatching.
Objective: minimize final makespan (Cmax) while respecting precedence and machine availability.
Static problem context:
Problem: 2 jobs x 2 machines (total_ops=4)
Fixed operation routes:
Job 0: ops=2, total_work=49, route=(M0,t30) -> (M1,t19)
Job 1: ops=2, total_work=... | Problem: 2 jobs x 2 machines (total_ops=4)
Fixed operation routes:
Job 0: ops=2, total_work=49, route=(M0,t30) -> (M1,t19)
Job 1: ops=2, total_work=40, route=(M1,t22) -> (M0,t18) | [
0,
1
] | [
"<A4453>",
"<A9997>"
] | {
"<A4453>": 0,
"<A9997>": 1
} | 1 | 0 | dispatch | true | jssp_step_reason_dispatch_v1_action_token | 1 | <A9997> | You are analyzing an already-selected JSSP action.
Objective: explain why this action was selected and why strong alternatives were not selected.
Ground the explanation in explicit evidence only: immediate timing/Cmax impact, bottleneck-machine use or release, downstream route exposure (next2), waiting/idle trade-offs,... | Reason: <A9997> is feasible on M0 (est_start=30, est_end=48). Projected Cmax changes 30->48 (delta=18). It completes this job, eliminating the remaining work from 18 to 0. It works on high-load machine M0 (remaining_load=18), which is close to the current bottleneck pressure. It intentionally waits 8 time units for M0,... | {
"action_code": "<A9997>",
"job_id": 1,
"machine_id": 0,
"machine_token": "M0",
"proc_time": 18,
"next_machine": 0,
"next_proc_time": 18,
"next2_machine": -1,
"next2_proc_time": 0,
"remaining_ops_before": 1,
"remaining_ops_after": 0,
"remaining_work_before": 18,
"remaining_work_after": 0,
"... | [
"<A4453>"
] | [
{
"action_code": "<A4453>",
"job_id": 0,
"machine_id": 1,
"machine_token": "M1",
"proc_time": 19,
"next_machine": 1,
"next_proc_time": 19,
"next2_machine": -1,
"next2_proc_time": 0,
"remaining_ops_before": 1,
"remaining_ops_after": 0,
"remaining_work_before": 19,
... | deterministic_teacher_v1 |
train_000003 | 3 | 2 | 2 | 4 | 3 | {
"step_idx": 3,
"total_steps": 4,
"scheduled_ratio": 0.75,
"current_time": 30,
"current_cmax": 48,
"job_next_op": [
1,
2
],
"job_total_ops": [
2,
2
],
"job_total_work": [
49,
40
],
"job_ready_time": [
30,
48
],
"machine_ready_time": [
48,
22
],
"n... | You are solving JSSP with event-driven dispatching.
Objective: minimize final makespan (Cmax) while respecting precedence and machine availability.
Static problem context:
Problem: 2 jobs x 2 machines (total_ops=4)
Fixed operation routes:
Job 0: ops=2, total_work=49, route=(M0,t30) -> (M1,t19)
Job 1: ops=2, total_work=... | Problem: 2 jobs x 2 machines (total_ops=4)
Fixed operation routes:
Job 0: ops=2, total_work=49, route=(M0,t30) -> (M1,t19)
Job 1: ops=2, total_work=40, route=(M1,t22) -> (M0,t18) | [
0
] | [
"<A5719>"
] | {
"<A5719>": 0
} | 1 | 1 | dispatch | true | jssp_step_reason_dispatch_v1_action_token | 0 | <A5719> | You are analyzing an already-selected JSSP action.
Objective: explain why this action was selected and why strong alternatives were not selected.
Ground the explanation in explicit evidence only: immediate timing/Cmax impact, bottleneck-machine use or release, downstream route exposure (next2), waiting/idle trade-offs,... | Reason: <A5719> is feasible on M1 (est_start=30, est_end=49). Projected Cmax changes 48->49 (delta=1). It completes this job, eliminating the remaining work from 19 to 0. It directly activates the current bottleneck machine M1 (remaining_load=19, ops_left=1), so delaying this move would postpone work on the heaviest un... | {
"action_code": "<A5719>",
"job_id": 0,
"machine_id": 1,
"machine_token": "M1",
"proc_time": 19,
"next_machine": 1,
"next_proc_time": 19,
"next2_machine": -1,
"next2_proc_time": 0,
"remaining_ops_before": 1,
"remaining_ops_after": 0,
"remaining_work_before": 19,
"remaining_work_after": 0,
"... | [] | [] | deterministic_teacher_v1 |
train_000004 | 4 | 2 | 2 | 4 | 0 | {
"step_idx": 0,
"total_steps": 4,
"scheduled_ratio": 0,
"current_time": 0,
"current_cmax": 0,
"job_next_op": [
0,
0
],
"job_total_ops": [
2,
2
],
"job_total_work": [
33,
35
],
"job_ready_time": [
0,
0
],
"machine_ready_time": [
0,
0
],
"next_machi... | You are solving JSSP with event-driven dispatching.
Objective: minimize final makespan (Cmax) while respecting precedence and machine availability.
Static problem context:
Problem: 2 jobs x 2 machines (total_ops=4)
Fixed operation routes:
Job 0: ops=2, total_work=33, route=(M1,t18) -> (M0,t15)
Job 1: ops=2, total_work=... | Problem: 2 jobs x 2 machines (total_ops=4)
Fixed operation routes:
Job 0: ops=2, total_work=33, route=(M1,t18) -> (M0,t15)
Job 1: ops=2, total_work=35, route=(M0,t17) -> (M1,t18) | [
0,
1
] | [
"<A4276>",
"<A8233>"
] | {
"<A4276>": 0,
"<A8233>": 1
} | 0 | 0 | dispatch | true | jssp_step_reason_dispatch_v1_action_token | 1 | <A8233> | You are analyzing an already-selected JSSP action.
Objective: explain why this action was selected and why strong alternatives were not selected.
Ground the explanation in explicit evidence only: immediate timing/Cmax impact, bottleneck-machine use or release, downstream route exposure (next2), waiting/idle trade-offs,... | Reason: <A8233> is feasible on M0 (est_start=0, est_end=17). Projected Cmax changes 0->17 (delta=17). It reduces this job from remaining_work 35->18 and remaining_ops 2->1, while exposing the next route on M1 (t=18). It also exposes a downstream step on bottleneck M1 (next2_p=18), pulling future critical work closer. I... | {
"action_code": "<A8233>",
"job_id": 1,
"machine_id": 0,
"machine_token": "M0",
"proc_time": 17,
"next_machine": 0,
"next_proc_time": 17,
"next2_machine": 1,
"next2_proc_time": 18,
"remaining_ops_before": 2,
"remaining_ops_after": 1,
"remaining_work_before": 35,
"remaining_work_after": 18,
... | [
"<A4276>"
] | [
{
"action_code": "<A4276>",
"job_id": 0,
"machine_id": 1,
"machine_token": "M1",
"proc_time": 18,
"next_machine": 1,
"next_proc_time": 18,
"next2_machine": 0,
"next2_proc_time": 15,
"remaining_ops_before": 2,
"remaining_ops_after": 1,
"remaining_work_before": 33,
... | deterministic_teacher_v1 |
train_000004 | 4 | 2 | 2 | 4 | 1 | {
"step_idx": 1,
"total_steps": 4,
"scheduled_ratio": 0.25,
"current_time": 0,
"current_cmax": 17,
"job_next_op": [
0,
1
],
"job_total_ops": [
2,
2
],
"job_total_work": [
33,
35
],
"job_ready_time": [
0,
17
],
"machine_ready_time": [
17,
0
],
"next... | You are solving JSSP with event-driven dispatching.
Objective: minimize final makespan (Cmax) while respecting precedence and machine availability.
Static problem context:
Problem: 2 jobs x 2 machines (total_ops=4)
Fixed operation routes:
Job 0: ops=2, total_work=33, route=(M1,t18) -> (M0,t15)
Job 1: ops=2, total_work=... | Problem: 2 jobs x 2 machines (total_ops=4)
Fixed operation routes:
Job 0: ops=2, total_work=33, route=(M1,t18) -> (M0,t15)
Job 1: ops=2, total_work=35, route=(M0,t17) -> (M1,t18) | [
0
] | [
"<A5514>"
] | {
"<A5514>": 0
} | 0 | 1 | dispatch | true | jssp_step_reason_dispatch_v1_action_token | 0 | <A5514> | You are analyzing an already-selected JSSP action.
Objective: explain why this action was selected and why strong alternatives were not selected.
Ground the explanation in explicit evidence only: immediate timing/Cmax impact, bottleneck-machine use or release, downstream route exposure (next2), waiting/idle trade-offs,... | Reason: <A5514> is feasible on M1 (est_start=0, est_end=18). Projected Cmax changes 17->18 (delta=1). It reduces this job from remaining_work 33->15 and remaining_ops 2->1, while exposing the next route on M0 (t=15). It directly activates the current bottleneck machine M1 (remaining_load=36, ops_left=2), so delaying th... | {
"action_code": "<A5514>",
"job_id": 0,
"machine_id": 1,
"machine_token": "M1",
"proc_time": 18,
"next_machine": 1,
"next_proc_time": 18,
"next2_machine": 0,
"next2_proc_time": 15,
"remaining_ops_before": 2,
"remaining_ops_after": 1,
"remaining_work_before": 33,
"remaining_work_after": 15,
... | [] | [] | deterministic_teacher_v1 |
train_000004 | 4 | 2 | 2 | 4 | 2 | {
"step_idx": 2,
"total_steps": 4,
"scheduled_ratio": 0.5,
"current_time": 18,
"current_cmax": 18,
"job_next_op": [
1,
1
],
"job_total_ops": [
2,
2
],
"job_total_work": [
33,
35
],
"job_ready_time": [
18,
17
],
"machine_ready_time": [
17,
18
],
"ne... | You are solving JSSP with event-driven dispatching.
Objective: minimize final makespan (Cmax) while respecting precedence and machine availability.
Static problem context:
Problem: 2 jobs x 2 machines (total_ops=4)
Fixed operation routes:
Job 0: ops=2, total_work=33, route=(M1,t18) -> (M0,t15)
Job 1: ops=2, total_work=... | Problem: 2 jobs x 2 machines (total_ops=4)
Fixed operation routes:
Job 0: ops=2, total_work=33, route=(M1,t18) -> (M0,t15)
Job 1: ops=2, total_work=35, route=(M0,t17) -> (M1,t18) | [
0,
1
] | [
"<A9488>",
"<A5395>"
] | {
"<A9488>": 1,
"<A5395>": 0
} | 1 | 0 | dispatch | true | jssp_step_reason_dispatch_v1_action_token | 0 | <A5395> | You are analyzing an already-selected JSSP action.
Objective: explain why this action was selected and why strong alternatives were not selected.
Ground the explanation in explicit evidence only: immediate timing/Cmax impact, bottleneck-machine use or release, downstream route exposure (next2), waiting/idle trade-offs,... | Reason: <A5395> is feasible on M0 (est_start=18, est_end=33). Projected Cmax changes 18->33 (delta=15). It completes this job, eliminating the remaining work from 15 to 0. It works on high-load machine M0 (remaining_load=15), which is close to the current bottleneck pressure. It fills machine idle gap=1 on M0, which he... | {
"action_code": "<A5395>",
"job_id": 0,
"machine_id": 0,
"machine_token": "M0",
"proc_time": 15,
"next_machine": 0,
"next_proc_time": 15,
"next2_machine": -1,
"next2_proc_time": 0,
"remaining_ops_before": 1,
"remaining_ops_after": 0,
"remaining_work_before": 15,
"remaining_work_after": 0,
"... | [
"<A9488>"
] | [
{
"action_code": "<A9488>",
"job_id": 1,
"machine_id": 1,
"machine_token": "M1",
"proc_time": 18,
"next_machine": 1,
"next_proc_time": 18,
"next2_machine": -1,
"next2_proc_time": 0,
"remaining_ops_before": 1,
"remaining_ops_after": 0,
"remaining_work_before": 18,
... | deterministic_teacher_v1 |
train_000004 | 4 | 2 | 2 | 4 | 3 | {
"step_idx": 3,
"total_steps": 4,
"scheduled_ratio": 0.75,
"current_time": 18,
"current_cmax": 33,
"job_next_op": [
2,
1
],
"job_total_ops": [
2,
2
],
"job_total_work": [
33,
35
],
"job_ready_time": [
33,
17
],
"machine_ready_time": [
33,
18
],
"n... | You are solving JSSP with event-driven dispatching.
Objective: minimize final makespan (Cmax) while respecting precedence and machine availability.
Static problem context:
Problem: 2 jobs x 2 machines (total_ops=4)
Fixed operation routes:
Job 0: ops=2, total_work=33, route=(M1,t18) -> (M0,t15)
Job 1: ops=2, total_work=... | Problem: 2 jobs x 2 machines (total_ops=4)
Fixed operation routes:
Job 0: ops=2, total_work=33, route=(M1,t18) -> (M0,t15)
Job 1: ops=2, total_work=35, route=(M0,t17) -> (M1,t18) | [
1
] | [
"<A1350>"
] | {
"<A1350>": 1
} | 1 | 1 | dispatch | true | jssp_step_reason_dispatch_v1_action_token | 1 | <A1350> | You are analyzing an already-selected JSSP action.
Objective: explain why this action was selected and why strong alternatives were not selected.
Ground the explanation in explicit evidence only: immediate timing/Cmax impact, bottleneck-machine use or release, downstream route exposure (next2), waiting/idle trade-offs,... | Reason: <A1350> is feasible on M1 (est_start=18, est_end=36). Projected Cmax changes 33->36 (delta=3). It completes this job, eliminating the remaining work from 18 to 0. It directly activates the current bottleneck machine M1 (remaining_load=18, ops_left=1), so delaying this move would postpone work on the heaviest un... | {
"action_code": "<A1350>",
"job_id": 1,
"machine_id": 1,
"machine_token": "M1",
"proc_time": 18,
"next_machine": 1,
"next_proc_time": 18,
"next2_machine": -1,
"next2_proc_time": 0,
"remaining_ops_before": 1,
"remaining_ops_after": 0,
"remaining_work_before": 18,
"remaining_work_after": 0,
"... | [] | [] | deterministic_teacher_v1 |
train_000005 | 5 | 2 | 2 | 4 | 0 | {
"step_idx": 0,
"total_steps": 4,
"scheduled_ratio": 0,
"current_time": 0,
"current_cmax": 0,
"job_next_op": [
0,
0
],
"job_total_ops": [
2,
2
],
"job_total_work": [
47,
53
],
"job_ready_time": [
0,
0
],
"machine_ready_time": [
0,
0
],
"next_machi... | You are solving JSSP with event-driven dispatching.
Objective: minimize final makespan (Cmax) while respecting precedence and machine availability.
Static problem context:
Problem: 2 jobs x 2 machines (total_ops=4)
Fixed operation routes:
Job 0: ops=2, total_work=47, route=(M0,t24) -> (M1,t23)
Job 1: ops=2, total_work=... | Problem: 2 jobs x 2 machines (total_ops=4)
Fixed operation routes:
Job 0: ops=2, total_work=47, route=(M0,t24) -> (M1,t23)
Job 1: ops=2, total_work=53, route=(M1,t26) -> (M0,t27) | [
0,
1
] | [
"<A2289>",
"<A1013>"
] | {
"<A2289>": 0,
"<A1013>": 1
} | 0 | 0 | dispatch | true | jssp_step_reason_dispatch_v1_action_token | 0 | <A2289> | You are analyzing an already-selected JSSP action.
Objective: explain why this action was selected and why strong alternatives were not selected.
Ground the explanation in explicit evidence only: immediate timing/Cmax impact, bottleneck-machine use or release, downstream route exposure (next2), waiting/idle trade-offs,... | Reason: <A2289> is feasible on M0 (est_start=0, est_end=24). Projected Cmax changes 0->24 (delta=24). It reduces this job from remaining_work 47->23 and remaining_ops 2->1, while exposing the next route on M1 (t=23). It directly activates the current bottleneck machine M0 (remaining_load=51, ops_left=2), so delaying th... | {
"action_code": "<A2289>",
"job_id": 0,
"machine_id": 0,
"machine_token": "M0",
"proc_time": 24,
"next_machine": 0,
"next_proc_time": 24,
"next2_machine": 1,
"next2_proc_time": 23,
"remaining_ops_before": 2,
"remaining_ops_after": 1,
"remaining_work_before": 47,
"remaining_work_after": 23,
... | [
"<A1013>"
] | [
{
"action_code": "<A1013>",
"job_id": 1,
"machine_id": 1,
"machine_token": "M1",
"proc_time": 26,
"next_machine": 1,
"next_proc_time": 26,
"next2_machine": 0,
"next2_proc_time": 27,
"remaining_ops_before": 2,
"remaining_ops_after": 1,
"remaining_work_before": 53,
... | deterministic_teacher_v1 |
train_000005 | 5 | 2 | 2 | 4 | 1 | {
"step_idx": 1,
"total_steps": 4,
"scheduled_ratio": 0.25,
"current_time": 0,
"current_cmax": 24,
"job_next_op": [
1,
0
],
"job_total_ops": [
2,
2
],
"job_total_work": [
47,
53
],
"job_ready_time": [
24,
0
],
"machine_ready_time": [
24,
0
],
"next... | You are solving JSSP with event-driven dispatching.
Objective: minimize final makespan (Cmax) while respecting precedence and machine availability.
Static problem context:
Problem: 2 jobs x 2 machines (total_ops=4)
Fixed operation routes:
Job 0: ops=2, total_work=47, route=(M0,t24) -> (M1,t23)
Job 1: ops=2, total_work=... | Problem: 2 jobs x 2 machines (total_ops=4)
Fixed operation routes:
Job 0: ops=2, total_work=47, route=(M0,t24) -> (M1,t23)
Job 1: ops=2, total_work=53, route=(M1,t26) -> (M0,t27) | [
1
] | [
"<A9856>"
] | {
"<A9856>": 1
} | 0 | 1 | dispatch | true | jssp_step_reason_dispatch_v1_action_token | 1 | <A9856> | You are analyzing an already-selected JSSP action.
Objective: explain why this action was selected and why strong alternatives were not selected.
Ground the explanation in explicit evidence only: immediate timing/Cmax impact, bottleneck-machine use or release, downstream route exposure (next2), waiting/idle trade-offs,... | Reason: <A9856> is feasible on M1 (est_start=0, est_end=26). Projected Cmax changes 24->26 (delta=2). It reduces this job from remaining_work 53->27 and remaining_ops 2->1, while exposing the next route on M0 (t=27). It directly activates the current bottleneck machine M1 (remaining_load=49, ops_left=2), so delaying th... | {
"action_code": "<A9856>",
"job_id": 1,
"machine_id": 1,
"machine_token": "M1",
"proc_time": 26,
"next_machine": 1,
"next_proc_time": 26,
"next2_machine": 0,
"next2_proc_time": 27,
"remaining_ops_before": 2,
"remaining_ops_after": 1,
"remaining_work_before": 53,
"remaining_work_after": 27,
... | [] | [] | deterministic_teacher_v1 |
train_000005 | 5 | 2 | 2 | 4 | 2 | {
"step_idx": 2,
"total_steps": 4,
"scheduled_ratio": 0.5,
"current_time": 26,
"current_cmax": 26,
"job_next_op": [
1,
1
],
"job_total_ops": [
2,
2
],
"job_total_work": [
47,
53
],
"job_ready_time": [
24,
26
],
"machine_ready_time": [
24,
26
],
"ne... | You are solving JSSP with event-driven dispatching.
Objective: minimize final makespan (Cmax) while respecting precedence and machine availability.
Static problem context:
Problem: 2 jobs x 2 machines (total_ops=4)
Fixed operation routes:
Job 0: ops=2, total_work=47, route=(M0,t24) -> (M1,t23)
Job 1: ops=2, total_work=... | Problem: 2 jobs x 2 machines (total_ops=4)
Fixed operation routes:
Job 0: ops=2, total_work=47, route=(M0,t24) -> (M1,t23)
Job 1: ops=2, total_work=53, route=(M1,t26) -> (M0,t27) | [
0,
1
] | [
"<A2040>",
"<A5044>"
] | {
"<A2040>": 1,
"<A5044>": 0
} | 1 | 0 | dispatch | true | jssp_step_reason_dispatch_v1_action_token | 1 | <A2040> | You are analyzing an already-selected JSSP action.
Objective: explain why this action was selected and why strong alternatives were not selected.
Ground the explanation in explicit evidence only: immediate timing/Cmax impact, bottleneck-machine use or release, downstream route exposure (next2), waiting/idle trade-offs,... | Reason: <A2040> is feasible on M0 (est_start=26, est_end=53). Projected Cmax changes 26->53 (delta=27). It completes this job, eliminating the remaining work from 27 to 0. It directly activates the current bottleneck machine M0 (remaining_load=27, ops_left=1), so delaying this move would postpone work on the heaviest u... | {
"action_code": "<A2040>",
"job_id": 1,
"machine_id": 0,
"machine_token": "M0",
"proc_time": 27,
"next_machine": 0,
"next_proc_time": 27,
"next2_machine": -1,
"next2_proc_time": 0,
"remaining_ops_before": 1,
"remaining_ops_after": 0,
"remaining_work_before": 27,
"remaining_work_after": 0,
"... | [
"<A5044>"
] | [
{
"action_code": "<A5044>",
"job_id": 0,
"machine_id": 1,
"machine_token": "M1",
"proc_time": 23,
"next_machine": 1,
"next_proc_time": 23,
"next2_machine": -1,
"next2_proc_time": 0,
"remaining_ops_before": 1,
"remaining_ops_after": 0,
"remaining_work_before": 23,
... | deterministic_teacher_v1 |
train_000005 | 5 | 2 | 2 | 4 | 3 | {
"step_idx": 3,
"total_steps": 4,
"scheduled_ratio": 0.75,
"current_time": 26,
"current_cmax": 53,
"job_next_op": [
1,
2
],
"job_total_ops": [
2,
2
],
"job_total_work": [
47,
53
],
"job_ready_time": [
24,
53
],
"machine_ready_time": [
53,
26
],
"n... | You are solving JSSP with event-driven dispatching.
Objective: minimize final makespan (Cmax) while respecting precedence and machine availability.
Static problem context:
Problem: 2 jobs x 2 machines (total_ops=4)
Fixed operation routes:
Job 0: ops=2, total_work=47, route=(M0,t24) -> (M1,t23)
Job 1: ops=2, total_work=... | Problem: 2 jobs x 2 machines (total_ops=4)
Fixed operation routes:
Job 0: ops=2, total_work=47, route=(M0,t24) -> (M1,t23)
Job 1: ops=2, total_work=53, route=(M1,t26) -> (M0,t27) | [
0
] | [
"<A6883>"
] | {
"<A6883>": 0
} | 1 | 1 | dispatch | true | jssp_step_reason_dispatch_v1_action_token | 0 | <A6883> | You are analyzing an already-selected JSSP action.
Objective: explain why this action was selected and why strong alternatives were not selected.
Ground the explanation in explicit evidence only: immediate timing/Cmax impact, bottleneck-machine use or release, downstream route exposure (next2), waiting/idle trade-offs,... | Reason: <A6883> is feasible on M1 (est_start=26, est_end=49). Projected Cmax changes 53->53 (delta=0). It completes this job, eliminating the remaining work from 23 to 0. It directly activates the current bottleneck machine M1 (remaining_load=23, ops_left=1), so delaying this move would postpone work on the heaviest un... | {
"action_code": "<A6883>",
"job_id": 0,
"machine_id": 1,
"machine_token": "M1",
"proc_time": 23,
"next_machine": 1,
"next_proc_time": 23,
"next2_machine": -1,
"next2_proc_time": 0,
"remaining_ops_before": 1,
"remaining_ops_after": 0,
"remaining_work_before": 23,
"remaining_work_after": 0,
"... | [] | [] | deterministic_teacher_v1 |
train_000006 | 6 | 2 | 2 | 4 | 0 | {
"step_idx": 0,
"total_steps": 4,
"scheduled_ratio": 0,
"current_time": 0,
"current_cmax": 0,
"job_next_op": [
0,
0
],
"job_total_ops": [
2,
2
],
"job_total_work": [
39,
46
],
"job_ready_time": [
0,
0
],
"machine_ready_time": [
0,
0
],
"next_machi... | You are solving JSSP with event-driven dispatching.
Objective: minimize final makespan (Cmax) while respecting precedence and machine availability.
Static problem context:
Problem: 2 jobs x 2 machines (total_ops=4)
Fixed operation routes:
Job 0: ops=2, total_work=39, route=(M1,t22) -> (M0,t17)
Job 1: ops=2, total_work=... | Problem: 2 jobs x 2 machines (total_ops=4)
Fixed operation routes:
Job 0: ops=2, total_work=39, route=(M1,t22) -> (M0,t17)
Job 1: ops=2, total_work=46, route=(M1,t23) -> (M0,t23) | [
0,
1
] | [
"<A2820>",
"<A0374>"
] | {
"<A2820>": 0,
"<A0374>": 1
} | 0 | 1 | dispatch | true | jssp_step_reason_dispatch_v1_action_token | 1 | <A0374> | You are analyzing an already-selected JSSP action.
Objective: explain why this action was selected and why strong alternatives were not selected.
Ground the explanation in explicit evidence only: immediate timing/Cmax impact, bottleneck-machine use or release, downstream route exposure (next2), waiting/idle trade-offs,... | Reason: <A0374> is feasible on M1 (est_start=0, est_end=23). Projected Cmax changes 0->23 (delta=23). It reduces this job from remaining_work 46->23 and remaining_ops 2->1, while exposing the next route on M0 (t=23). It directly activates the current bottleneck machine M1 (remaining_load=45, ops_left=2), so delaying th... | {
"action_code": "<A0374>",
"job_id": 1,
"machine_id": 1,
"machine_token": "M1",
"proc_time": 23,
"next_machine": 1,
"next_proc_time": 23,
"next2_machine": 0,
"next2_proc_time": 23,
"remaining_ops_before": 2,
"remaining_ops_after": 1,
"remaining_work_before": 46,
"remaining_work_after": 23,
... | [
"<A2820>"
] | [
{
"action_code": "<A2820>",
"job_id": 0,
"machine_id": 1,
"machine_token": "M1",
"proc_time": 22,
"next_machine": 1,
"next_proc_time": 22,
"next2_machine": 0,
"next2_proc_time": 17,
"remaining_ops_before": 2,
"remaining_ops_after": 1,
"remaining_work_before": 39,
... | deterministic_teacher_v1 |
End of preview. Expand in Data Studio
No dataset card yet
- Downloads last month
- 1